Editor

Assistant Editor

Contributors

Keo McKenzie, Mercedes Subhani, Avi Toltzis, and Alex Crowley

§ 1.1. Introduction

Another year has passed, and the legal issues arising out of Artificial Intelligence’s increasingly broad adoption across every facet of our lives continued to expand. As business lawyers, there is optimism that AI (especially Generative and Agentic AI) will create new opportunities for us to assist our clients in this rapidly developing legal landscape of lawsuits, proposed and newly enacted regulations, and novel IP, employment and privacy concerns. At the same time, lawyers must recognize how AI is changing the practice of law and clients’ expectations of how their outside counsel will work efficiently and smartly to further their interests.

Simply stated, lawyers have an ethical obligation to learn, understand and keep apace with how AI can be used in the daily practice of law. And as I write this, the media reports Bill Gates is predicting that in 10 years, most jobs across industries will become obsolete thanks to AI; something that has concerned me since at least 2014 (for those interested in my perspective, which remains largely unchanged 11 years later, please see my 2014 Tech Crunch article on the subject). Whether or not Mr. Gates is correct (and I would wager that he is), there is no denying AI’s impact on the legal profession and the seismic shift upon us in the way we utilize Generative AI to perform tasks which hitherto were the sole provenance of paralegals, legal assistants and junior lawyers. Tasks like legal research, document automation, risk and compliance management, and many other legal functions are rapidly becoming the provenance of AI tools and AI-powered legal assistants.

The pace of change with regard to AI’s use in the legal profession will only increase in the coming years, changing nearly everything about how our data-driven profession operates (“find a case that holds Y”; “pull up the last contract and compare it with the draft opposing counsel just sent,” “what is the law in X jurisdiction on Y issue”; “for an opening statement, pull the stats on how many drunk driving collisions end in fatalities,” “analyze this draft crypto loan agreement and suggest ways to strengthen it,” “find how many times Person Y is mentioned in all of the various FTX related filings around the country,” “for the 2 TB data set, find all of the documents that reference ‘Z,’ explain the new EU regulations on [subject] and our compliance obligations,” etc.). Like our clients, it is our duty to utilize AI in responsible and transparent ways.

For 2024, we have continued our practice of focusing on cases decided in 2024 and legislation enacted in 2024. Not surprisingly, emerging themes for both the courts and state and local legislators center around copyright infringement, privacy, fairness/perceived bias, civil rights, transparency and consent. The headline is that for the foreseeable future, practicing law will increasingly mean staying abreast of AI technology in many different use cases and domains. Our clients are looking for ways to produce and utilize AI to make themselves more efficient while cutting costs and errors associated with human capital. In the near future, it is likely that clients and judges will expect lawyers to utilize AI to better serve them and the courts. And like it or not, AI will remain the focus for regulators and litigants across the United States.

We hope that this Chapter continues to be a useful tool for lawyers looking for a straightforward summary of the major AI cases and legislation for 2024. And my colleagues Adam Aft, Keo McKenzie, Mercedes Subhani, Avi Toltzis, and Alex Crowley have my gratitude for their assistance in preparing this year’s Chapter.

We look forward to tracking the trends in these cases and presenting the cases arising over the next several years.

Bradford Newman

Editor and Co-Chair of the AI and Blockchain Subcommittee of the Business and Corporate Litigation Committee

Palo Alto

§ 1.2. Artificial Intelligence Cases of Note

§ 1.2.0. United States Supreme Court

Moody v. NetChoice, LLC, 603 U.S. 707, 144 S. Ct. 2383 (2024). Defendant NetChoice, an internet trade association, alleged that Florida and Texas laws restricting social media platforms’ ability to moderate content on their websites via algorithms (including artificial intelligence) violated the First Amendment to the U.S. Constitution. The Court found that the lower courts had not properly considered the issues and vacated and remanded the prior judgments regarding each law.

§ 1.2.1. First Circuit

Harris v. Adams, No. 24-cv-12437-PGL, 2024 U.S. Dist. LEXIS 210951 (D. Mass. Nov. 20, 2024). Two high school students were punished for cheating on an AP U.S. History project by failing to attribute the source for text (which included hallucinations) that they copied from Grammarly’s artificial intelligence software. Plaintiff Harris claimed that the high school violated Plaintiff’s due process rights and that the punishments were too harsh. The Court denied Plaintiff’s motion for preliminary injunction because Plaintiff had failed to show any misconduct by authorities of Plaintiff’s schools.

Overjet, Inc. v. VideaHealth, Inc., Civil Action No. 24-cv-10446-ADB, 2024 U.S. Dist. LEXIS 128030 (D. Mass. July 19, 2024). Overjet and VideaHealth compete in providing artificial intelligence-enabled dental software. In this case, Overjet alleged that Videa infringed Overjet’s copyrights related to its software and falsely advertised, including regarding Videa’s software’s artificial intelligence capabilities, in violation of the Lanham Act. The Court denied Overjet’s motion for a preliminary injunction because Overjet had not sufficiently shown a likelihood that its claims would succeed on the merits nor a likelihood of irreparable harm.

WEX Inc. v. HP Inc., No. 2:24-cv-00121-JAW, 2024 U.S. Dist. LEXIS 119715 (D. Me. July 9, 2024). WEX alleged that HP’s “HP WEX” software name infringed WEX’s trademark “WEX.” HP argued that a news article presented as evidence of confusion of the WEX mark and HP WEX “‘was created using generative artificial intelligence,’ and therefore ‘no person . . . was confused.’” WEX countered that the article was nonetheless reviewed by a confused human editor. The Court found that the AI-generated article indicated that HP and WEX could be confused as being affiliated given the “HP WEX” brand. The Court granted WEX’s motion for a preliminary injunction against HP regarding use of “WEX.”

Baker v. CVS Health Corp., 717 F. Supp. 3d 188 (D. Mass. Feb. 16, 2024). A job candidate alleged that CVS violated the Massachusetts Lie Detector Statute (Mass. Gen. Laws. Ch. 149, §19B) by subjecting the candidate to an artificial intelligence-based test (to help evaluate an individual’s integrity and cultural fit) during a job interview without notifying the candidate of his statutory rights. The Court denied CVS’ motions to dismiss for failure to state a claim and for lack of standing.

§ 1.2.2. Second Circuit

N.Y. Times Co. v. Microsoft Corp., No. 23-cv-11195 (SHS) (OTW), 2024 U.S. Dist. LEXIS 212998 (S.D.N.Y. Nov. 22, 2024). The Court denied Open AI’s motion to compel production of disputed discovery because Open AI failed to demonstrate the relevance of asking for the New York Times’ documents related to the Times’ use of non-parties’ generative AI tools due to it neither being relevant nor proportional to Open AI’s fair use defense.

Dukuray v. Experian Info. Sols., 2024 U.S. Dist. LEXIS 132667 (S.D.N.Y. July 26, 2024). The Court did not believe any sanctions would be appropriate against a pro se Plaintiff because they would not be aware of the risk that ChatGPT and similar AI programs can generate fake case citations and other misstatements of law in their Fair Credit Reporting Act case.

Z.H. v. N.Y.C. Dep’t of Educ., No. 23-cv-3081 (ER), 2024 U.S. Dist. LEXIS 124478 (S.D.N.Y. July 12, 2024). The Judge declined to credit evidence the Firm submitted using ChatGPT that showed the Firm’s requested rates are reasonable market rates because ChatGPT has been shown to be an unreliable source.

Gross v. Madison Square Garden Ent. Corp., No. 23-cv-3380 (LAK) (JLC), 2024 U.S. Dist. LEXIS 83102 (S.D.N.Y. May 7, 2024). The Court granted the defendant’s motion to dismiss the complaint for failure to state a claim because it held that sharing biometric data with a third party to implement a policy banning certain individuals from venues does not constitute “profiting” from the biometric data within the meaning of NYC Ad. Code § 22-1202(b).

Network-1 Techs., Inc. v. Google LLC & YouTube, LLC, 2024 U.S. Dist. LEXIS 76545 (S.D.N.Y. Apr. 24, 2024). The Court held that in this patent infringement case against Google, the term “non-exhaustive search” is indefinite because persons skilled in the art could reasonably construe it in different ways based on the intrinsic and extrinsic evidence. Therefore, even though Google’s Siberia version of Content ID conducted an algorithmic patent search which increased computing resources, it did not perform a sublinear search as required by the ’237 Patent because undisputed evidence showed the search to be linear and the Plaintiff failed to show the multi-step search as a whole is sublinear.

Rensselaer Polytechnic Inst. v. Amazon.Com, Inc., 723 F. Supp. 3d 132 (N.D.N.Y. Mar. 18, 2024). The Court denied Plaintiffs’ motion for summary judgement and granted Defendant’s motion for summary judgement due to its ’798 patent of an approach for interpreting and responding to a natural language input by storing and searching certain types of information not being subject matter eligible for patent protection under 35 U.S.C. § 101.

Park v. Kim, 91 F.4th 610 (2d Cir. Jan. 30, 2024). The Court referred an attorney to the Second Circuit’s Grievance Panel for investigation for submitting a brief that relief on “non-existent” caselaw generated by ChatGPT.

§ 1.2.3. Third Circuit

Thomson Reuters Enter. Ctr. GmbH v. Ross Intel. Inc., No. 1:20-cv-613-SB, 2025 U.S. Dist. LEXIS 24296 (D. Del. Feb. 11, 2025) and Thomson Reuters Enter. Ctr. GmbH v. Ross Intel. Inc., No. 1:20-cv-613-SB, 2024 U.S. Dist. LEXIS 175507 (D. Del. Sep. 27, 2024). Ross Intelligence aimed to improve legal research via development of an artificial intelligence-based research tool. Thomson Reuters alleged that Ross infringed Thomson Reuters’ copyrighted Westlaw headnotes and Key Number System by using them in the development of Ross’s tool. In its February 2025 opinion, the Court “grant[ed] most of Thomson Reuters’s motion for partial summary judgment on direct copyright infringement and related defenses, D.I. 674; (2) grant[ed] Thomson Reuters’s motion for partial summary judgment on fair use, D.I. 672; (3) den[ied] Ross’s motion for summary judgment on fair use, D.I. 676; and (4) den[ied] Ross’s motion for summary judgment on Thomson Reuters’s copyright claims, D.I. 683.”

In parallel to Thomson Reuters’s copyright infringement claim, Ross alleged that Thomson Reuters’s Westlaw caselaw database and search tools were tied together in violation of antitrust laws. The Court rejected Ross’s allegation and granted summary judgment to Thomson Reuters with respect to Ross’s antitrust claims.

Huckabee v. Meta Platforms, Inc., Civil Action No. 24-773-GBW, 2024 U.S. Dist. LEXIS 209624 (D. Del. Nov. 18, 2024). Former Governor of Arkansas Mike Huckabee alleged that Meta should be liable under various laws and for violation of various legal rights for allowing, via its machine learning algorithms, presentation of third-party advertisements that made false claims and falsely attributed statements to Governor Huckabee. The Court held that Meta was liable under Section 230 of the Communications Decency Act as an “information content provider” because its algorithms determined advertisement presentation, but the Court denied Governor Huckabee’s claims as they did not state claims upon which relief can be granted.

VB Assets, LLC v. Amazon.com Servs. LLC, No. 19-1410 (MN), 2024 U.S. Dist. LEXIS 176993 (D. Del. Sep. 30, 2024). Plaintiff VB Assets alleged that Amazon’s Alexa voice assistant and associated devices violated VB Assets’ smart speaker technology patents. The Court granted Amazon’s motion for judgment as a matter of law for only one of VB Assets’ infringement claims.

Lee v. ElectrifAI, LLC, Civil Action No. 23-2239 (JXN) (JRA), 2024 U.S. Dist. LEXIS 165093 (D.N.J. Sep. 13, 2024). As part of this case, the Court rejected plaintiff’s claim that her prior employer, ElectrifAI, LLC was misrepresenting the functionality of artificial intelligence in its products. The Court found that the plaintiff did not provide sufficient facts to establish a plausible claim.

Elkin Valley Baptist Church v. PNC Bank, N.A., Civil Action No. 23-1798, 2024 U.S. Dist. LEXIS 162888 (W.D. Pa. Sep. 10, 2024). In this case regarding statutory interpretation of Section 4A-207 (pertains to financial fraud) of the Uniform Commercial Code, the Court notes that financial institutions’ use of artificial intelligence and other automation necessitates developing a well-reasoned interpretation of the Section.

State Farm Mut. Auto. Ins. Co. v. Amazon.Com, Inc., Civil Action No. 22-1447-CJB, 2024 U.S. Dist. LEXIS 160437 (D. Del. Sep. 6, 2024). State Farm alleged that Amazon infringed several of State Farm’s patents relating to use of machine learning and neural networks to evaluate whether an individual can safely live independently. In this case, the Court denied Amazon’s motion to dismiss under which Amazon argued that the asserted patents pertained to patent ineligible subject matter and were thus invalid.

IPA Techs. Inc. v. Microsoft Corp., Civil Action No. 18-1-RGA, 2024 U.S. Dist. LEXIS 76038 (D. Del. Apr. 25, 2024). IPA Technologies alleged that Microsoft products containing Microsoft’s virtual assistant Cortana infringed on various of IPA Technologies’ patents regarding software architecture that “supports cooperative task completion by flexible and autonomous electronic agents.” The Court granted in part, denied in part, and dismissed as moot in part Plaintiff’s and Defendant’s summary judgment and Daubert motions.

§ 1.2.4. Fourth Circuit

Saas v. Major, Lindsey & Africa, LLC, No. 1:23-cv-02102-JRR, 2024 U.S. Dist. LEXIS 84968 (D. Md. May 10, 2024). Plaintiff Saas alleged that the defendants used algorithmic and machine learning tools in their recruitment processes, which led to unlawful discrimination based on sex and age in violation of Title VII of the Civil Rights Act and the Age Discrimination in Employment Act of 1967. Saas claimed that these tools discriminated against women with employment gaps due to motherhood, which caused them to be passed over for interviews and other opportunities. However, the plaintiff’s claims that the defendant used discriminatory AI tools was based solely on the supposition that all large businesses use AI tools, which was contradicted by the defendant’s statement that it does not use AI tools. The Court therefore concluded that the plaintiff had not adequately pleaded discrimination through the use of AI tools. The decision highlights the importance for plaintiffs to plead specific facts relating to a defendant’s use of AI tools in respect of discrimination claims.

§ 1.2.5. Fifth Circuit

Mullen Indus. LLC v. Meta Platforms, Inc., No. 1:24-CV-00354-DAE, 2024 U.S. Dist. LEXIS 207934 (W.D. Tex. Nov. 14, 2024). Mullen Industries alleged that Meta’s augmented and virtual reality systems infringed on twelve of its patents, including a claim on the use of AI technologies. The Magistrate Judge for this case recommended that the District Court grant Meta’s motion to dismiss in respect of the AI claims, because it found that Mullen had not plausibly pled that AI was present in the allegedly infringing systems.

Hicks v. Collier, No. 2:24-CV-00126, 2024 U.S. Dist. LEXIS 241129 (S.D. Tex. Oct. 31, 2024). Plaintiff, a Texas prisoner, alleged that his constitutional rights were violated due to excessively hot living conditions and inadequate medical care. A significant aspect of the case involved the Texas Department of Criminal Justice’s (“TDCJ”) use of an algorithm to classify inmates for housing assignments based on their “heat scores.” Hicks claimed that this algorithm misclassified him, leading to his placement in non-air-conditioned housing, which exacerbated his health issues. The Court’s decision to preserve the claims against the TDCJ and its officials highlights the need for transparency, accuracy, and accountability in AI implementation to protect individual rights. However, the Court dismissed the plaintiff’s claims under § 1983 against the unidentified developer of the algorithm because the Court found that it was not acting under the color of state law.

§ 1.2.6. Sixth Circuit

Concord Music Grp., Inc. v. Anthropic PBC, 738 F. Supp. 3d 973 (M.D. Tenn. 2024). Several music publishers sued Anthropic, an AI research company, alleging that Anthropic used their copyrighted song lyrics to train its AI model, Claude, without proper authorization. The Court found that it lacked personal jurisdiction over Anthropic, a Delaware company with its principal place of business in California (and which used data located in Virginia to train the Claude model, which itself was hosted on servers in Iowa), and transferred the action to California. The Court rejected plaintiffs’ arguments that Anthropic had availed itself of sufficient contacts with the forum state by making the model available to Tennessee through an interactive website. This decision will be significant insofar as it rejects the notion that personal jurisdiction can be established over the developer of an AI model simply by the developer’s making that model available in the jurisdiction.

§ 1.2.7. Seventh Circuit

G.T. v. Samsung Elecs. Am., Inc., No. 21 CV 4976, 2024 U.S. Dist. LEXIS 233003 (N.D. Ill. Dec. 23, 2024) and G.T. v. Samsung Elecs. Am. Inc., No. 21 CV 4976, 2024 U.S. Dist. LEXIS 130771 (N.D. Ill. July 24, 2024). In a first amended complaint, plaintiffs in this class action lawsuit alleged that Samsung violated Illinois’s Biometric Information Privacy Act (“BIPA”) in possessing and collecting their biometric data via Samsung’s Gallery photo applications. The Court found claims to be insufficiently pled and granted Samsung’s motion to dismiss.

In their second amended complaint, the plaintiffs alleged that Samsung violated BIPA by offering software that generates and stores biometric data (face templates) on a device using facial recognition technology. The Court granted Samsung’s second motion to dismiss on the basis that BIPA requires control over the actual biometric data, and Samsung did not have such control.

Arnold v. Target Corp., No. 24 CV 4452, 2024 U.S. Dist. LEXIS 212009 (N.D. Ill. Nov. 21, 2024). Plaintiffs alleged that Target Corp. violated BIPA in possessing, collecting, and disclosing their biometric data (face geometry captured by facial recognition technology in Target stores). The Court denied Target’s motion to dismiss on the basis that plaintiffs’ claims were plausible.

Hartman v. Meta Platforms, Inc., No. 3:23-CV-02995-NJR, 2024 U.S. Dist. LEXIS 167696 (S.D. Ill. Sep. 17, 2024). Plaintiffs in this putative class action lawsuit alleged that Meta violated BIPA in possessing and collecting their biometric data via augmented reality features of the Facebook Messenger and Messenger Kids applications. The Court denied Meta’s motion to dismiss on the basis that plaintiffs’ claims were plausible, and the case proceeded to discovery.

Lewerentz v. 1411 State Parkway Condo. Ass’n, No. 23-cv-1635, 2024 U.S. Dist. LEXIS 159664 (N.D. Ill. Sep. 5, 2024). A building engineer continued to receive calls from elevator call buttons after stopping work at those buildings. The calls included an artificial intelligence voice. The engineer alleged that the calls were harassment under the Telephone Consumer Protection Act and a tort of intrusion upon seclusion under Illinois state law. The Court found that these claims were insufficiently pled and granted the Defendant’s motion to dismiss the complaint.

Plumbers v. Morris Plumbing, LLC, No. 23-CV-616-JPS-JPS, 2024 U.S. Dist. LEXIS 70751 (E.D. Wis. Apr. 18, 2024). In this case, the Court noted that a case in Plaintiff’s reply brief appeared to be hallucinated by artificial intelligence, as the case could not be found via online searching. The Court warned Plaintiff’s counsel that they would be sanctioned for any future presentations of non-existent cases.

Taylor v. 48forty Sols., LLC, No. 23 C 14400, 2024 U.S. Dist. LEXIS 64573 (N.D. Ill. Apr. 9, 2024). In this putative class action, a truck driver alleged that his former employer collected scans of his face geometry (biometric data) in violation of BIPA Sections 15(a), (b), and (d). The Court denied the former employer’s motion to dismiss all those claims, though plaintiff was required to be ready to inform the Court whether he would like to proceed with the Section 15(a) claim in federal or state court.

Hernandez v. Omnitracs, LLC, No. 1:22-CV-00109, 2024 U.S. Dist. LEXIS 58865 (N.D. Ill. Mar. 31, 2024). In this putative class action, a truck driver alleged that his former employer collected scans of his face geometry (biometric data) in violation of BIPA Sections 15(a)-(d). The Court denied the former employer’s motion to dismiss all those claims.

§ 1.2.8. Eighth Circuit

No cases identified for the Eighth Circuit.

§ 1.2.9. Ninth Circuit

Tate v. VITAS Healthcare Corp., No. 2:24-cv-01327-DJC-CSK, 2025 U.S. Dist. LEXIS 3828 (E.D. Cal. Jan. 8, 2025). VITAS uses third party conversation intelligence software, records calls, creates transcripts, and uses AI to classify data into a searchable database. Plaintiff, who interacted with the software to discuss hospice care for her mother, alleged violations of California Invasion of Privacy Act (CIPA). The Court found that the AI software could be considered a third party and a recording device under CIPA and denied VITAS’s motion to dismiss.

Netchoice v. Bonta, No. 5:24-cv-07885-EJD, 2024 U.S. Dist. LEXIS 234919 (N.D. Cal. Dec. 31, 2024). The Court considered the constitutionality of SB 976, the Protecting Our Kids from Social Media Addiction Act, to regulate social media platforms’ interactions with minors. Among other requirements, SB 976 restricts personalized feeds and notifications for minors. The Court found that algorithms designed to maximize a person’s time spent on social media do not reflect any message from its creator and therefore do not constitute expressive speech. The Court found that the plaintiff had not met its burden of establishing that the personalized feed provisions of the law impermissibly restrict free speech and dismissed those elements of the claim.

Ryan v. X Corp., No. 24-cv-03553-WHO, 2024 U.S. Dist. LEXIS 222459 (N.D. Cal. Dec. 9, 2024). Plaintiff Ryan alleges X Corp. used AI to target and suspend his accounts without proper notice. The Court granted X Corp.’s motion to dismiss after finding that all claims were barred by X Corp.’s Terms of Service, which limit liability for account suspensions. Additionally, the unjust enrichment claim was also barred by Section 230 of the Communications Decency Act and X Corp.’s use of AI to moderate content does not negate Section 230 immunity. Ryan was given leave to amend his complaint.

Vance v. Google LLC, No. 20-cv-04696-BLF, 2024 U.S. Dist. LEXIS 220639 (N.D. Cal. Dec. 5, 2024). In this case, the Court denied the motion to dismiss the plaintiffs’ claims under section 15(b) of Illinois’s BIPA but granted dismissal of their Section 15(c) claim. Plaintiffs posted photos containing their faces to Flickr, a photo hosting website. IBM created the Diversity in Faces (DiF) Dataset using Flickr photos without user permission and Google obtained the DiF Dataset from IBM to improve facial recognition technology for its facial unlock feature. The Court found that improving a product was not sufficient to demonstrate a commercial transaction to support a Section 15(c) claim.

Samuels v. Dao, No. 23-cv-06492-VC, 2024 U.S. Dist. LEXIS 209474 (N.D. Cal. Nov. 18, 2024). This case was brought by an investor who bought cryptocurrency tokens issued by Lido DAO and lost money on his investment. The Court rejected defendant’s argument that it is merely autonomous software that runs without human management and therefore not a legal entity that can be subject to legal proceedings. Rather, the alleged actions are of an entity run by people, and this entity can be sued as a general partnership.

Kohls v. Bonta, No. 2:24-cv-02527 JAM-CKD, 2024 U.S. Dist. LEXIS 179933 (E.D. Cal. Oct. 2, 2024). The Court granted a preliminary injunction against California’s AB 2839, which aims to address the spread of AI-generated “deepfakes” and other manipulated media that could mislead voters or undermine confidence in the electoral process. The Court found the law unconstitutional for being overly broad and not narrowly tailored, thus violating the First Amendment. The decision highlighted the role of AI in creating “deepfakes” and emphasized the importance of protecting free speech involving digitally manipulated content.

Lamontagne v. Tesla, Inc., No. 23-cv-00869-AMO, 2024 U.S. Dist. LEXIS 178030 (N.D. Cal. Sep. 30, 2024). The plaintiffs alleged that Tesla, Inc., and Elon Musk made twenty-nine false or misleading statements about the development and safety of Tesla’s autonomous driving technology. The Court granted Tesla’s motion to dismiss, finding that the statements were either protected by the PSLRA safe harbor, nonactionable corporate puffery, or not sufficiently alleged to be false or misleading. The Court also dismissed the plaintiffs’ claims under Sections 10(b) and 20(a) of the Securities Exchange Act, as well as claims under Items 105 and 303 of Regulation S-K, due to insufficient allegations of scienter and materiality. The plaintiffs were given leave to amend their complaint.

Andersen v. Stability AI Ltd., No. 23-cv-00201-WHO, 2024 U.S. Dist. LEXIS 143204 (N.D. Cal. Aug. 12, 2024). The Court partially granted and partially denied the defendants’ motions to dismiss. The plaintiffs, a group of artists, alleged that Stability AI and other defendants used their copyrighted works to train AI models without permission. The Court allowed the copyright infringement claims to proceed, finding the plaintiffs’ allegations plausible. However, it dismissed the Digital Millennium Copyright Act (“DMCA”) claims and unjust enrichment claims. The Court found that the plaintiffs’ infringement allegations regarding the use of their works in training AI models were sufficiently pled. The court granted the plaintiffs leave to amend their unjust enrichment claims complaint but dismissed the DMCA claims with prejudice.

Mobley v. Workday, Inc., No. 23-cv-00770-RFL, 2024 U.S. Dist. LEXIS 126336 (N.D. Cal. July 12, 2024). The Court granted in part and denied in part the defendant’s motion to dismiss an employment discrimination case. The plaintiff, Derek Mobley, alleged that Workday’s algorithm-based applicant screening tools discriminated against him and others based on race, age, and disability. The Court denied the claims under Title VII, ADEA, and ADA based on Workday’s liability as an agent of employers. However, it granted the claims based on Workday being an employment agency and intentional discrimination claims under Title VII, ADEA, and Section 1981. Additionally, the Court granted with leave to amend the claims under California’s Fair Employment and Housing Act (FEHA). Mobley was permitted twenty-one days to amend his complaint regarding the FEHA claim.

Jones v. Peloton Interactive, Inc., No. 23-cv-1082-L-BGS, 2024 U.S. Dist. LEXIS 118511 (S.D. Cal. July 5, 2024). The Court denied Peloton’s motion to dismiss the First Amended Complaint. The plaintiffs claimed that Peloton violated the California Invasion of Privacy Act (CIPA) by using AI-powered third-party software, Drift, to intercept and record chat communications on its website without users’ consent. The Court found that Drift’s AI technology, which analyzed and used the intercepted data for its own purposes, acted as a third-party eavesdropper. Accordingly the plaintiffs’ CIPA claims were sufficiently pled.

Ambriz v. Google, LLC, No. 23-cv-05437-RFL, 2024 U.S. Dist. LEXIS 119619 (N.D. Cal. June 20, 2024). The Court dismissed Misael Ambriz’s complaint against Google, which alleged that Google’s Cloud Contact Center AI wiretapped, eavesdropped on, and recorded his call to Verizon’s customer service. The Court found that Google’s AI acted as a virtual agent for Verizon, a telephone company, and thus fell under the exemption provided by the California Invasion of Privacy Act (CIPA).

Forrest v. Meta Platforms, Inc., No. 22-cv-03699-PCP, 2024 U.S. Dist. LEXIS 107340 (N.D. Cal. June 17, 2024). The Court partially granted and partially denied Meta’s motion to dismiss. The case centered on Facebook advertisements using Dr. Andrew Forrest’s likeness to promote fraudulent investments. The advertisements were alleged to have been created and optimized by Meta’s AI and machine learning tools, raising factual disputes about Meta’s liability under Section 230 that were unsuitable for preliminary resolution. Specifically the pleadings left a factual dispute as to whether Meta materially contributes to the ads to render it beyond Section 230’s protection. The claims for misappropriation and negligence were allowed to proceed.

Dental Monitoring SAS v. Align Tech., Inc., No. C 22-07335 WHA, 2024 U.S. Dist. LEXIS 88739 (N.D. Cal. May 16, 2024). The Court granted the defendant’s summary judgment motion, invalidating Dental Monitoring’s patents, which involved methods for remote dental aligner assessment using deep learning devices. Applying the Alice two-step test, the Court determined that the invention was both directed to an abstract concept and that it merely applied generic machine learning technology to the known field of dental aligners.

Alich v. Opendoor Techs. Inc., No. CV-22-01717-PHX-MTL, 2024 U.S. Dist. LEXIS 86544 (D. Ariz. May 14, 2024). The Court granted the plaintiffs’ motion for reconsideration of the Court’s earlier dismissal, thereby allowing claims under Sections 11 and 15 of the Securities Act of 1933 to proceed. The plaintiffs alleged that Opendoor made misleading statements about its algorithm’s ability to adjust to market conditions, which they claimed led to investors’ financial losses. The defendant’s algorithm was designed to adjust dynamically to market indicators and economic conditions. The Court found that misrepresentations regarding Opendoor’s algorithm touches upon the alleged reasons for plaintiffs’ losses, finding that the plaintiffs have adequately pleaded a Section 11 claim.

Gibson v. Cendyn Grp., LLC, No. 2:23-cv-00140-MMD-DJA, 2024 U.S. Dist. LEXIS 83547 (D. Nev. May 8, 2024). The Court dismissed the plaintiffs’ claims with prejudice. The plaintiffs had alleged that the defendants, including a software company and several hotel operators, violated the Sherman Antitrust Act by artificially inflating hotel room prices through the use of pricing algorithms. The Court found that the plaintiffs failed to plausibly allege a tacit agreement among the defendants to fix prices and that the vertical agreements between the software company and the hotel operators did not restrain trade. The Court found that the mere use of algorithmic pricing, without allegations of any explicit or implicit agreement between competitors to accept the prices that the algorithm recommends, does not enable a plausible allegation of illegal collusion.

Tremblay v. OpenAI, Inc., 716 F. Supp. 3d 772 (N.D. Cal. Feb. 12, 2024). The Court granted in part and denied in part OpenAI’s motions to dismiss. The plaintiffs, authors of copyrighted books, alleged that OpenAI used their works to train its language models without permission. The Court dismissed claims for vicarious copyright infringement, DMCA violations, negligence, and unjust enrichment, but allowed the unfair competition claim to proceed. The plaintiffs were permitted to amend their complaint with respect to the dismissed claims.

Meta Platforms, Inc. v. Bright Data Ltd., No. 23-cv-00077-EMC, 2024 U.S. Dist. LEXIS 11913 (N.D. Cal. Jan. 23, 2024). The Court granted the defendant, Bright Data’s motion to dismiss finding that its scraping of publicly available data while not logged into a user account did not breach Meta’s Terms of Service. The Court also determined that the Terms did not apply to Bright Data’s activities after it terminated its accounts, and the survival clause did not impose a perpetual ban on scraping public data. This led to the dismissal of Meta’s breach of contract claims.

§ 1.2.10. Tenth Circuit

United States v. Cole, No. 1:24-cr-00054-SKC, 2024 U.S. Dist. LEXIS 184877 (D. Colo. Oct. 8, 2024). Criminal defendant Cole argued that the “unique selection” of his image using facial recognition software contributed to making his image stand out in a photo array. The Court did not find that such selection caused the array to be impermissibly suggestive.

MarketDial, Inc. v. Applied Predictive Techs., Inc., No. 1:23-cv-00477-JNP-CMR, 2024 U.S. Dist. LEXIS 109809 (D. Utah June 20, 2024). MarketDial, Inc. alleged that Applied Predictive Technologies, Inc.’s (APT) patent directed toward “determining optimal parameter settings for a predictive machine-learning model in business initiative testing software” was invalid or unenforceable. The Court determined that the patent failed the Alice test for patent eligibility, granted MarketDial’s motion to dismiss APT’s counterclaim of patent infringement, and denied APT’s motion to dismiss the complaint.

Total Quality Sys. v. Universal Synaptics Corp., No. 1:22-cv-00167-RJS-DAO, 2024 U.S. Dist. LEXIS 93224 (D. Utah May 23, 2024). In this case, Universal Synaptics Corporation alleged that Total Quality Systems infringed two of Universal’s patents, one of which covers an apparatus containing a neural network. Applying the Alice test, the Court held that the claimed inventions were ineligible for patent protection under 35 U.S.C. § 101.

§ 1.2.11. Eleventh Circuit

United States v. Deleon, 116 F.4th 1260 (11th Cir. 2024) and Snell v. United Specialty Ins. Co., 102 F.4th 1208 (11th Cir. 2024). In two cases, Judge Kevin C. Newsom of the United States Court of Appeals, Eleventh Circuit, wrote concurring opinions in which he evaluates how AI-based large language models could aid in conducting interpretive analysis in line with an “ordinary meaning” approach to evaluating legal texts.

Mazile v. Larkin Univ. Corp., No. 1:23-cv-23306-LEIBOWITZ, 2024 U.S. Dist. LEXIS 128457 (S.D. Fla. July 22, 2024). Larkin University expelled student/plaintiff Mazile after an AI system owned by remote testing company ExamSoft flagged that Mazile had cheated on a test monitored by the AI system. Mazile brought claims against ExamSoft and Larkin. The Court granted ExamSoft’s motion to compel arbitration under the End User License Agreement to which Mazile had agreed. The Court dismissed Mazile’s claim for discrimination based on her disability because Mazile failed to provide evidence that Larkin knew that ExamSoft’s AI system was discriminatory, and that Larkin discriminated against Mazile because of her disability.

Medallia Inc. v. Echospan, Inc., No. 1:23-cv-3730-TCB, 2024 U.S. Dist. LEXIS 160154 (N.D. Ga. June 14, 2024). Medallia Inc. asserted that Echospan, Inc. infringed Medallia’s patent regarding sentiment analysis of text. The Court denied Echospan’s motion asserting that Medallia’s patent was directed to patent-ineligible subject matter because the parties had not yet agreed on the meaning of critical terms “first model” and “relevantly similar analysis model” in the patent claims.

Doe v. Emory Univ., 734 F. Supp. 3d 1369 (N.D. Ga. 2024). Two students at Emory University created an “artificial intelligence-based learning tool” that Emory’s Honor Council determined may be used for cheating. Emory initiated disciplinary proceedings against the students. In this case, the Court rejects one student’s motion to proceed in litigation anonymously because the student did not satisfy precedential requirements for permitting anonymity in court, despite the student’s assertions that they could be subject to negative attention if their identity was made public.

§ 1.2.12. DC Circuit

TikTok Inc. & ByteDance Ltd. v. Garland, 122 F.4th 930 (D.C. Cir. 2024). The federal Protecting Americans from Foreign Adversary Controlled Applications Act, enacted in April 2024, results in the ban of Tik-Tok’s AI-enabled social media app in the US. Among other claims, TikTok asserted that the Act violated freedom of speech under the First Amendment. The Court rejected the TikTok’s First Amendment claim on the basis that the Act’s provisions addressed compelling national security interests “to counter (1) the PRC’s efforts to collect data of and about persons in the United States, and (2) the risk of the PRC covertly manipulating content on TikTok.” The U.S. Supreme Court affirmed this judgment in January 2025. See TikTok Inc. v. Garland, 145 S. Ct. 57 (2025).

Rubio v. District of Columbia, Civil Action No. 23-719 (RDM), 2024 U.S. Dist. LEXIS 218004 (D.D.C. Dec. 3, 2024). In this case, the Court denied all the plaintiff’s federal and D.C. law claims, including based on the plaintiff’s provision of cases likely fabricated by AI.

Biddle v. DOD, Civil Action No. 23-1380 (TJK), 2024 U.S. Dist. LEXIS 164961 (D.D.C. Sep. 13, 2024). Plaintiff Biddle requested certain “records pertaining to the Algorithmic Warfare Cross-Functional Team’s use of Google technology, software or hardware” from the Department of Defense via a Freedom of Information Act (FOIA) request. The Department asserted that disclosure of its approach to AI development and implementation in response to the FOIA request would “reveal vulnerabilities in Department of Defense critical infrastructure.” The Court was unconvinced, including because an approach to AI is not clearly “infrastructure.” The Court denied both parties’ motions for summary judgment.

United States v. Google LLC, 747 F. Supp. 3d 1 (D.D.C. 2024). As part of a broader case alleging that Google was engaged in monopolistic practices in violation of antitrust law, Google asserted that the rapid development of AI eroded barriers to entry to providing general search services. The Court rejected that assertion on the basis that AI had not yet developed sufficiently to “change the market dynamic in the ‘foreseeable future’.”

§ 1.2.13. Court of Appeals for the Federal Circuit

Promptu Sys. Corp. v. Comcast Corp., 92 F.4th 1372 (Fed. Cir. 2024). Plaintiff Promptu Systems alleged that Comcast infringed its patents related to speech or voice recognition technology. The Federal Circuit Court of Appeals vacated the district court’s judgment with respect to certain claims and remanded the case for further proceedings.

§ 1.3. Legislation

As in 2023, legislation governing the development, deployment, and use of artificial intelligence continued to be a hot topic in 2024. Below, we summarize key substantive artificial intelligence legislation enacted in 2024.

§ 1.3.0. Multiple States

Deepfakes and sexual offenses. Many states enacted laws in 2024 related to deepfakes and sexual offenses. We list those states and their laws below.

• Alabama:
  • H.B. 168, Alabama Child Protection Act of 2024
• Delaware:
  • H.B. 353, An Act to Amend Titles 10 and 11 of the Delaware Code Relating to Deep Fakes
• Florida:
  • S.B. 1680, Advanced Technology
• Idaho:
  • H.B. 465, An Act Relating to Crimes Against Children
  • H.B. 575, An Act Relating to Disclosing Explicit Synthetic Media
• Indiana:
  • H.B. 1047, Sexual Offenses
• Louisiana:
  • S.B. 6, An Act to enact R.S. 14:73.14, relative to computer related crime; to create the crime of unlawful dissemination or sale of images of another created by artificial intelligence; to provide definitions; to provide penalties; and to provide for related matters
• Pennsylvania:
  • S.B. 1213, An Act amending Titles 18 (Crimes and Offenses) and 61 (Prisons and Parole) of the Pennsylvania Consolidated Statutes, in sexual offenses, further providing for the offense of unlawful dissemination of intimate image; in minors, further providing for the offense of sexual abuse of children and for the offense of transmission of sexually explicit images by minor; and making editorial changes to replace references to the term “child pornography” with references to the term “child sexual abuse material”
• Tennessee:
  • H.B. 2163, An Act to amend Tennessee Code Annotated, Title 39 and Title 40, relative to the sexual exploitation of children
• Washington:
  • H.B. 1999, Concerning fabricated intimate or sexually explicit images and depictions

Deepfakes and election protection. Many states enacted laws in 2024 related to deepfakes and election protection. We list those states and their laws below.

• Alabama:
  • H.B. 172, Relating to elections; to provide that distrib. of materially deceptive media is a crime
• Arizona:
  • S.B. 1359, Election communications; deepfakes; prohibition
  • H.B. 2394, Digital impersonation; injunctive relief; requirements
• California:
  • A.B. 2355, Political Reform Act of 1974: political advertisements: artificial intelligence
  • A.B. 2655, Defending Democracy from Deepfake Deception Act of 2024
  • A.B. 2839, Elections: deceptive media in advertisements
    • Note that enforcement of A.B. 2839 was partially enjoined under Kohls v. Bonta, No. 2:24-cv-02527 JAM-CKD, 2024 U.S. Dist. LEXIS 179933, as described above.
• Colorado:
  • H.B. 24-1147, Candidate Election Deepfake Disclosures
• Delaware:
  • H.B. 316, An Act to Amend Title 15 of the Delaware Code Relating to Deep Fakes in Elections
• Florida:
  • H.B. 919, Artificial Intelligence Use in Political Advertising
• Hawaii:
  • S.B. 2687, Elections; Materially Deceptive Media; Artificial Intelligence; Deepfake Technology; Prohibition; Penalty; Remedies
• Minnesota:
  • H.F. 4772, Elections policy and finance bill
• Mississippi:
  • S.B. 2577, An Act to Create a New Section in Title 97, Chapter 13, Mississippi Code of 1972, to Create Criminal Penalties for the Wrongful Dissemination of Digitizations; and for related purposes
• New Hampshire:
  • H.B. 1596, An Act requiring a disclosure of deceptive artificial intelligence usage in political advertising
  • H.B. 1432, An Act relative to prohibiting certain uses of deepfakes and creating a private claim of action
• New Mexico:
  • H.B. 182, An Act relating to Elections; amending and enacting sections of the Campaign Reporting Act by adding disclaimer requirements for advertisements containing materially deceptive media; creating the crime of distributing or entering into an agreement with another person to distribute materially deceptive media; adding definitions; providing penalties
• Oregon:
  • S.B. 1571, Relating to the use of artificial intelligence in campaign communications; declaring an emergency
• Utah:
  • S.B. 131 (includes clauses touching on elections and criminal justice)
• Wisconsin:
  • A.B. 664, An Act to amend 11.1303 (title); and to create 11.1303 (2m) of the statutes; relating to: disclosures regarding content generated by artificial intelligence in political advertisements, granting rule-making authority, and providing a penalty

§ 1.3.1. California

A.B. 1008, California Consumer Privacy Act of 2018: personal information. Enacted in September 2024, this act revises the scope of personal information under California’s Consumer Privacy Act to allow for multiple formats in which personal information may exist, including abstract digital formats such as “artificial intelligence systems that are capable of outputting personal information.”

A.B. 1836, Use of likeness: digital replica. Enacted in September 2024, this act creates a cause of action for damages when a digital replica of a deceased person is used without prior consent from the person’s estate.

A.B. 2013, Generative artificial intelligence: training data transparency. Enacted in September 2024, this act requires developers of generative AI systems released to California residents on or after January 1, 2022, to disclose details about the data used to train the systems, including a summary of the relevant datasets.

A.B. 2602, Contracts against public policy: personal or professional services: digital replicas. Enacted in September 2024, this act makes unenforceable certain contract provisions regarding performance of services by a digital replica of an individual in lieu of that individual’s own work on or after January 1, 2025.

A.B. 2905, Telecommunications: automatic dialing-announcing devices: artificial voices. Enacted in September 2024, this act amends California’s requirements regarding automated phone calls to further require notification to the call recipient if a prerecorded message uses an artificial voice (generated or significantly altered via artificial intelligence).

A.B. 2885, Artificial Intelligence. Enacted in September 2024, this act amends various sections of California state law to define artificial intelligence as “an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments.”

S.B. 942, California AI Transparency Act. Enacted in September 2024, this act applies to producers of generative artificial intelligence systems that have over one million monthly users and are publicly accessible within California. The producers must comply with various transparency requirements, such as providing free AI detection tools, and enabling disclosure of and disclosing when content is generated by AI.

A.B. 3030, Health care services: artificial intelligence. Enacted in September 2024, this act requires health care providers to notify patients when communication is performed using generative AI.

S.B. 1120, Health care coverage: utilization review. Enacted in September 2024, this act imposes various requirements on use of AI by health care service plan or disability insurers in performing utilization review or utilization management functions.

§ 1.3.2. Colorado

S.B. 24-205, Consumer Protections in Interactions with Artificial Intelligence Systems. Enacted in May 2024, this act requires developers and deployers of high-risk artificial intelligence systems to use reasonable care to protect Colorado consumers from any known or reasonably foreseeable risks of algorithmic discrimination.

§ 1.3.3. Illinois

H.B. 3773, Limit Predictive Analytics Use. Enacted in August 2024, this act amends Illinois’s list of civil rights violations to include (1) use of artificial intelligence for employment decision purposes that subjects employees to discrimination based on a protected class or zip code as proxy for a protected class, and (2) failure to notify employees of use of artificial intelligence for employment decision purposes.

H.B. 4875, Publicity Act—Use of AI. Enacted in August 2024, this act provides artists with rights to control use of digital replicas of them.

H.B. 4762, Digital Voice and Likeness Protection Act. Enacted in August 2024, this act imposes various requirements on contractual negotiations intended to permit creation and use of digital replicas of an individual.

§ 1.3.4. New Hampshire

H.B. 1688, An Act relative to use of artificial intelligence by state agencies. Enacted in May 2024, this act restricts state agencies from discriminating against people using AI, using AI for biometric surveillance, and using deepfakes for deceptive or malicious purposes.

§ 1.3.5. New York

S. 9832, New York State Fashion Workers Act. Enacted in December 2024, this act requires model management companies to obtain clear written consent to create or use, or alter or modify using artificial intelligence, a model’s digital replica.

S. 7676B, Establishes contract requirements for contracts involving the creation and use of digital replicas. Enacted in December 2024, this act imposes various requirements on contractual negotiations intended to permit creation and use of digital replicas of an individual.

S. 7543A, Enacts the legislative oversight of automated decision-making in government act (LOADinG Act). Enacted in December 2024, this act imposes various requirements, including disclosure requirements, on state agency use of automated decision-making systems.

§ 1.3.6. Tennessee

S.B. 1711, An Act to amend Tennessee Code Annotated, Title 49, relative to artificial intelligence. Enacted in March 2024, this act requires Tennessee state universities and public schools “to adopt a policy regarding the use of artificial intelligence by students, faculty, and staff for instructional and assignment purposes.”

H.B. 2091, Ensuring Likeness, Voice, and Image Security Act of 2024 (ELVIS Act). Enacted in March 2024, this act expanded existing law granting a property right in a person’s name, photograph, or likeness to include a property right in the person’s voice (including a simulation of the voice).

§ 1.3.7. Utah

S.B. 149, Artificial Intelligence Policy Act. Enacted in March 2024, this act requires disclosure of the provision of generative artificial intelligence-enabled services to a user, including for services of a regulated occupation. The act also establishes various initiatives related to artificial intelligence in Utah.

H.B. 366. Enacted in 2024, this act limits how an algorithm or risk assessment tool score may be used in various criminal justice procedures.

Shortly before the scheduled start of trial, the U.S. Department of Justice (“DOJ”), Antitrust Division (“Division”) reached a settlement with Hewlett Packard Enterprise (“HPE”) and Juniper Networks (“Juniper”) that allows their $14 billion merger to proceed. The settlement, described by the agency as “novel,” requires divestiture of an HPE business line to a preapproved buyer and at least one license of certain Juniper technology to one or more licensees that must be approved by the Division.

For the third time in a month, the new administration has approved a structural remedy in order to address the potential anticompetitive effects of a merger.

The Transaction

HPE offers products in a number of technology markets, including general-purpose servers, cloud storage, and finance. The company also sells networking products, including wireless access points and campus switches, under the HPE Aruba Networking brand and its legacy on-premises network management solution, Airwave. Juniper provides a range of networking products, including wireless access points, wired switches, and network management software under the Mist brand.

After the merger, HPE and Juniper’s aggregate market share would be only approximately 22–26 percent, below the 2023 Merger Guidelines’ 30 percent market share threshold for presumption of a merger’s illegality. However, the Division also alleged that the parties’ largest competitor has an approximate 48 percent market share and that at least seven other competitors each have market shares of only between 1 percent and 10 percent for commercial or enterprise-grade wireless networking solutions. The transaction would result in two firms controlling over 70 percent of the relevant market, with a significant gap between post-closing HPE and the next largest competitor in the market, allegedly making it easier for the two largest companies to reach and sustain a consensus on price, features, and reliability.

Though the transaction was cleared by fourteen foreign antitrust authorities, the Division sued to block the merger in January 2025 over concerns about competition for local wireless networking technology. According to the agency’s complaint, there were three primary theories of harm: (1) loss of head-to-head competition between the merging parties’ Aruba and Mist brands, causing prices to increase; (2) elimination of a disruptive force in the industry that has introduced tools to significantly lower the cost of wireless networks; and (3) increased risk of coordination among the remaining vendors.

The European Commission’s public findings regarding the transaction’s impact in the European Economic Area are in stark contrast with the Division’s allegations. Recent statements by HPE’s CEO might, however, explain the divergence; he has said that the transaction would facilitate the firm’s ability to better compete outside the United States, where more competitors with higher market shares participate in the market.

The Remedies

The divestiture and technology license(s) required by the settlement are intended to eliminate the alleged anticompetitive effects of the acquisition by strengthening one or more existing competitors or facilitating entry of a new competitor for enterprise-grade wireless local area network (“WLAN”) solutions.

• HPE must divest its global “Instant On” campus and branch WLAN business, including all assets, intellectual property, R&D personnel, and customer relationships within 180 days.
• The parties must also hold an auction for a perpetual, worldwide, nonexclusive license to Juniper’s AI Ops for Mist source code. The license will include optional transitional support “on reasonable commercial terms” and personnel transfers.

The settlement also assures that any winning licensee will have the right to any improvements to and derivatives of the licensed technology and the right to grant rights of use to the technology to its end users and service providers as reasonably needed. If the auction results in multiple bids exceeding $8 million, Juniper will be required to license to at least one additional bidder. This novel approach by the Justice Department reflects a commitment to solving unique challenges in mergers.

While not routine, license remedies have been used previously. For example, in 2017, the Federal Trade Commission (“FTC”) accepted a license remedy for it challenge to a pharmaceutical company’s acquisition of the US rights to the drug Synacthen. The FTC alleged there that the acquisition would prevent the development of a US competitor to the buyer’s monopoly. In another instance, a licensing remedy was approved in a post-consummation merger challenge.

A licensing remedy alone, however, would likely have been insufficient here. The divestiture of a business is an important component to the settlement with HPE and Juniper. Parties considering transactions should not assume that a license alone will resolve agency concerns. According to public reporting, the settlement was not supported by Division staff but was instead approved by leadership of the DOJ. Even assuming an antitrust enforcement divide within the administration, it is at least clear that there is a willingness to resolve merger challenges in advance of trial as was the case here and in advance of complaint in two other recent transactions. This shift in approach should be taken into consideration when assessing the enforcement risk of potential transactions, when designing agency clearance strategies, and when negotiating antitrust risk-shifting provisions in purchase agreements.

In a significant victory for environmental advocacy groups, the U.S. Supreme Court’s refusal on June 30, 2025, to grant certiorari in two pivotal cases—Port of Tacoma v. Puget Soundkeeper Alliance^[1] and ExxonMobil Corp. v. Environment Texas Citizen Lobby^[2]—is widely expected to embolden and increase the prevalence of citizen suits under federal environmental statutes. The Court’s decision leaves intact lower-court rulings that affirm a broad scope for citizen enforcement, reinforcing the role of individuals and organizations as “private attorneys general” in holding polluters accountable.

The denial of certiorari in these cases sends a clear message: The current legal framework empowering citizens to enforce environmental laws remains robust. Industry petitioners in both cases had sought to significantly narrow the reach of citizen suit provisions, particularly concerning standing requirements and the ability to enforce state-issued permits that go beyond federal standards. The Court’s inaction signals a rejection of these attempts to curb environmental watchdog efforts.

Port of Tacoma: Upholding State Permit Enforcement

In Port of Tacoma, the petitioners challenged a U.S. Court of Appeals for the Ninth Circuit ruling that allowed environmental groups to enforce state-issued Clean Water Act (“CWA”) permits in federal court, even when those permits contained requirements more stringent than federal law.^[3] This case centered on alleged violations of Washington State’s Industrial Stormwater General Permit, with Puget Soundkeeper Alliance arguing for accountability for polluted runoff from a wharf.

The Supreme Court’s denial of cert in this instance means that the Ninth Circuit’s expansive interpretation stands. This outcome is crucial because it allows citizens in circuits aligned with the Ninth (and Fourth and Eleventh) Circuit to continue to enforce the full scope of state-issued National Pollutant Discharge Elimination System (“NPDES”) permits. This is a considerable win for states seeking to implement stricter environmental protections and for citizen groups dedicated to upholding them.

It also highlights a continuing circuit split on this issue, as the U.S. Court of Appeals for the Second Circuit has a narrower view. But for now, the broader interpretation prevails in a significant portion of the country.

ExxonMobil: Affirming Broad Standing and Penalties

The ExxonMobil case involved a prolonged battle over air pollution from ExxonMobil’s Baytown, Texas, petrochemical complex. The U.S. Court of Appeals for the Fifth Circuit had upheld a substantial $14.25 million civil penalty against Exxon, the largest ever in a Clean Air Act (“CAA”) citizen suit.^[4] ExxonMobil had urged the Supreme Court to revisit its 2000 precedent in Friends of the Earth, Inc. v. Laidlaw Environmental Services (TOC), Inc., which established that civil penalties paid to the government could satisfy Article III’s “redressability” requirement for citizen plaintiffs.^[5] The company also sought to impose a more restrictive standard for demonstrating “traceability” of injuries to specific violations.^[6]

By denying cert, the Supreme Court left the Fifth Circuit’s decision undisturbed, affirming the broad interpretation of standing for citizen plaintiffs and reiterating that civil penalties can serve as a deterrent and redress injuries in citizen suits. This outcome is a significant affirmation of the fundamental principles underpinning environmental citizen enforcement, ensuring that groups like Environment Texas and the Sierra Club can continue to pursue accountability for environmental violations and secure substantial penalties that discourage future noncompliance.

The Landscape Ahead: More Citizen Suits Expected

The implications of these denials are clear: Environmental groups are poised to leverage these victories to intensify their citizen suit efforts. The CWA and CAA, along with over a dozen other federal environmental statutes, explicitly include citizen suit provisions, recognizing that government enforcement agencies may not always have the resources or the political will to pursue every violation.^[7]

As federal enforcement priorities potentially shift, citizen suits are expected to become an even more vital backstop, filling any perceived gaps in regulatory oversight. This means regulated entities should anticipate a heightened risk of litigation from environmental organizations, underscoring the critical importance of robust environmental compliance programs. The ability of citizens to seek both injunctive relief and significant civil penalties payable to the U.S. Treasury provides a powerful incentive for companies to adhere to environmental regulations.

In essence, the Supreme Court’s recent order list has reinforced the power of the public in environmental protection, solidifying the legal avenues for citizens to act as guardians of the nation’s air and water. This decision ensures that citizen suits will continue to be a formidable force in environmental litigation for the foreseeable future.

What This Means for Industry

1. Increased Legal Risk: Industries should anticipate a rise in citizen-led litigation, particularly from well-organized advocacy groups capable of using regulatory data and compliance reports to build strong cases.
2. Compliance Pressure: Regulatory compliance will no longer be a shield only scrutinized by the government. Any violation, even minor or self-reported, may trigger a lawsuit from a citizen group.
3. Litigation Costs: Even unsuccessful suits can entail significant legal costs and reputational damage. For example, ExxonMobil’s litigation spanned over a decade, involving appeals and millions in legal fees.

How Companies Can Protect Themselves

To mitigate this new landscape of liability, companies—especially those operating in manufacturing, energy, and logistics—should adopt proactive risk-management strategies.

1. Robust Compliance Systems: Implement and document environmental controls in real time. Use third-party audits to verify compliance with discharge, emissions, and permit conditions.
2. Transparency and Community Engagement: Foster open communication with local communities and environmental groups. Transparency can reduce adversarial relationships and build trust.
3. Rapid Incident Response: Ensure that any environmental release, even if minor, is addressed immediately and logged properly. Self-correction and voluntary reporting may reduce the likelihood of a lawsuit.
4. Legal Risk Assessments: Conduct periodic assessments to identify areas where past or current practices may be vulnerable to citizen suits. Remediate proactively.

Looking Ahead

The Supreme Court’s June 30 decision sends a clear signal: Citizen suits are here to stay—and may even be growing in power and frequency. While this strengthens environmental accountability, it also places added pressure on industries to go beyond minimum legal compliance and embrace more holistic environmental governance practices.

As the regulatory landscape evolves, businesses must prepare not only to meet federal standards but also to defend their records in the court of public and legal opinion.

1. Port of Tacoma, No. 24-350, 2025 WL 1787738 (U.S. June 30, 2025) (denial of cert). ↑

2. ExxonMobil, No. 24-982 (U.S. June 30, 2025) (denial of cert). ↑

3. Port of Tacoma, 104 F.4th 95 (9th Cir. 2024); see also 33 U.S.C. § 1365. ↑

4. ExxonMobil, 968 F.3d 357 (5th Cir. 2020); see also 42 U.S.C. § 7604. ↑

5. Friends of the Earth, 528 U.S. 167, 187 (2000). ↑

6. See ExxonMobil, 968 F.3d at 368. ↑

7. 33 U.S.C. § 1365; 42 U.S.C. § 7604. ↑

The Mendes Hershman Student Writing Contest is a highly regarded legal writing competition that encourages and rewards law students for their outstanding writing on business law topics. Papers are judged on research and analysis, choice of topic, writing style, originality, and contribution to the literature available on the topic. The distinguished former Business Law Section Chair Mendes Hershman (1974–1975) lends his name to this legacy. Read the abstract of this year’s second-place winner, Jake Granese of University of Miami School of Law, Class of 2026, below. Visit the University of Miami Business Law Review website to read the full article, published in Volume 34.

Sports betting in the United States has exploded in recent years, with a record-breaking $11 billion in revenue in 2023.^[1] This growth has inspired state governments to try to leverage this lucrative business. One recent development is the 2021 Gaming Compact between the Seminole Tribe of Florida and the state of Florida.^[2] The Compact represents a milestone towards exclusivity in mobile gaming for Native American tribal nations. Under the Compact, the Seminole Tribe receives sports betting exclusivity, for both retail sports wagering and online sports wagering, all across Florida, in exchange for a revenue sharing arrangement—unlike any other form of exclusivity. In the landmark case West Flagler Associates v. Haaland, the D.C. Circuit Court of Appeals upheld the Compact, ruling that any bet made off tribal land, but processed through servers located on tribal land, can be legally considered within tribal jurisdiction.^[3] This case solidified the Compact’s legality after a prolonged legal battle.

This ruling paves the way for sports betting legalization in other states with Native American tribal nations, shifting the balance of tribal-state relations throughout the country. Although the economic rewards of such arrangements are attractive, there are serious secondary concerns. One main worry is monopolistic control, detrimental to consumers through limited competition, suppressed market innovation, and increased costs. Though tribal sovereignty is important, policymakers need to also consider consumer protections as a top priority. Without proper regulation and oversight, sports betting offerings could yield predatory practices that exploit vulnerable customers. As more states move to legalize sports betting, regulators must guarantee competitive and fair markets for all consumers.

1. Marcus Lu, Visualizing the Growth of U.S. Sports Betting, Visual Capitalist (July 5, 2024). ↑

2. See generally 2021 Gaming Compact Between the Seminole Tribe of Florida and the State of Florida, Fla.–Seminole Tribe of Florida (Apr. 23, 2021) (available at Letter from Bryan Newland, Principal Deputy Assistant Sec’y – Indian Affs., U.S. Dep’t of the Interior, to the Hon. Marcellus W. Osceola, Jr., Chairman, Seminole Tribe of Florida (Aug. 6, 2021), at 13). ↑

3. W. Flagler Assocs., Ltd. v. Haaland, 71 F.4th 1059, 1066 (D.C. Cir. 2023). ↑

The continued expansion of e-commerce has driven a number of recent trends in intellectual property (“IP”) enforcement that require businesses to reexamine their IP assets’ form and protection. These include a shift in enforcement from traditional, registered IP to unregistered IP; an increase in IP enforcement against gray market goods; and an increase in infringement claims against online marketplaces. Policy changes such as new and increased tariffs may also affect IP enforcement. The effect of each trend will differ depending on the nature of a business, but business counsel should prioritize addressing them, as IP assets increasingly impact a business’s competitive edge and inform its future partnerships.

A Shift Toward Enforcement of Unregistered IP

Alice Corporation v. CLS Bank International, a landmark 2014 Supreme Court decision, generally weakened patent rights by invalidating a patent whose central concept was an abstract idea that merely required generic computer implementation and failed to sufficiently transform the abstract idea. Since then, many more patents, particularly in the software industry, have been challenged. In addition, the patent application process may take several years, and protection from a granted patent lasts only twenty years. This can lock businesses into IP that may or may not be valuable in the next decade, depending on e-commerce fads and new technologies. Because virality is often the source of a product’s success, flexibility in the ability to enforce different types of IP may be the most valuable commodity.

Accordingly, in addition to maintaining registered IP assets (like patents), businesses may benefit from developing a robust trade secret program and investing in unregistered trade dress and trademark protection. Counsel should help businesses increase their portfolio of unregistered IP, which offer flexibility to pursue infringement claims on an ad hoc basis. Unlike a patent or a registered trade dress or trademark, unregistered IP is amorphous—allowing the IP owner the flexibility to define its assets at the time the need to litigate against a competitor arises. For example, rather than have potential infringement allegations constrained by the four corners of a design patent, a business can define its trade dress as the exact elements of its own product packaging that are similar to its competitor’s packaging.

Strategies Against Gray Market Goods

There has been an increase in IP enforcement against resellers (often independent sellers) of gray market goods on e-commerce platforms because of the threat gray market goods pose to businesses’ revenue and brand integrity. Gray market goods are genuine goods originating from a brand owner but distributed outside of authorized channels. Some argue that distribution of gray market goods undercuts a brand’s control over its products. They further argue that these goods confuse customers, as listings of the same item on different platforms may offer different prices, with the cheaper gray market good lacking other official features like warranty policies or packaging. Accordingly, some believe this discrepancy hurts the brand’s reputation for consistency, reliable quality inspection, and transparent pricing, all of which ultimately devalues the brand’s IP.

For brand owners, trademark infringement claims are available to prevent sales of gray market goods. Brand owners typically bring a trademark infringement claim under Section 32 of the Lanham Act for registered marks and Section 43(a) for unregistered marks. An initial defense to these claims is the first sale doctrine, which protects subsequent lawful resellers of a product from trademark infringement claims because trademark owners no longer have the right to control distribution after the first sale—as long as the product is not materially altered and the source of the product is clear. To defeat this defense, brand owners should add intangibles to their products sold on authorized channels. Classic examples include warranty and satisfaction guarantee programs, but newer ideas could include access to online apps and profiles, exclusively for those who purchase the product through authorized channels.

Counsel for resellers should take note that brand owners have started pursuing false advertising claims under Section 43(a) of the Lanham Act. These claims are typically based on a reseller’s improper description of the product—for example, selling a used product as new. Claims for inadequate packaging or delay in delivery are also possible. The evidence for these claims is often gathered from feedback provided on the reseller’s account. To avoid false advertising claims, businesses (especially smaller shops) should clearly and unambiguously describe their product.

Policing Online Platforms

Policing IP infringement on e-commerce platforms is relatively simple for trademarks but becomes more complicated when the IP asset at issue is trade dress or a patented design. Keyword searches, which can be useful for identifying trademark infringement, are usually inadequate to hunt for and locate products that infringe these other IP rights.

In response to such difficulties, some brand owners bring infringement claims against the platform that hosts the infringing products rather than against the sellers (often foreign counterfeiters). To avoid liability and comply with the Digital Millennium Copyright Act, however, online marketplaces need only have a robust program where they take down infringing products upon notice and remove repeat infringers—and they typically do.

Though take-down programs are largely sufficient to meet their legal obligations, online marketplaces may find a proactive monitoring system a worthwhile strategy to consider and implement. With the sheer number of products listed on online marketplaces, traditional forms of monitoring are no longer capable of efficiently identifying infringing postings. As such, marketplaces should consider using artificial intelligence tools to monitor their websites for counterfeiting. These tools can identify patterns of counterfeiting within large numbers of listings and proactively take down infringing products even before receiving a notice.

The Impact of Tariffs—Counterfeit Goods and Licensing Agreement Disruption

At a surface level, tariffs are taxes imposed on foreign suppliers that subsequently impact business operations and supply chains in the U.S. and decrease the profits of those who export to the U.S. However, tariffs also have far-reaching ramifications for intellectual property that businesses should be aware of to stay ahead of the curve.

Businesses that source materials from foreign outlets face increased costs of operation and strained relationships with long-term suppliers because of tariffs. Two results tend to follow: (1) an influx of counterfeit goods trying to capitalize on the market’s dissatisfaction with the inflated price of legitimate goods and (2) a disruption in licensing as tariffs affect pricing and royalty structures and consequently cross-border licensing agreements.

Counsel and businesses should consider the following actions in light of these developments. First, if production is moved to a different jurisdiction in response to a shift in trade agreements, counsel should conduct an IP audit in the new jurisdiction to examine risks and ensure the enforceability of any IP assets. Second, to combat counterfeit products, tools like unique serial numbers, digital watermarking, and blockchain-based product tracking should be considered to bolster protection. Third, clauses in agreements including royalty adjustments, exclusivity terms, and termination triggers should be revisited in light of the tariffs.

Because the creation, acquisition and protection of IP is an investment and discretionary spend, tariffs may disincentivize the procurement of U.S. IP by foreign suppliers. However, because tariffs can make U.S. companies more profitable, it is possible that with increased tariffs, U.S. companies will enhance their investment in IP assets. In a tense competitive environment, foreign suppliers may resort to litigation to regain their declined competitive standing, while U.S. companies may be feeling less litigious. Counsel for U.S. businesses should be aware of this litigation potential and proactively analyze what IP protection the business has in its arsenal should the need to bring a claim against a competitor arise.

In light of the risks posed by the digital marketplace and the ever-evolving economic landscape, it is imperative that businesses and their legal advisors remain up to date with IP developments and consider incorporating flexible unregistered IP, product-associated intangibles, and artificial intelligence monitoring systems to strengthen their IP protection while reevaluating their supply chains and operations.

There are few issues as sensitive in private mergers and acquisitions (“M&A”) as “sandbagging.” As deal lawyers know well, “sandbagging” refers to a scenario where an M&A buyer brings a post-closing indemnification claim based on a breached seller representation and warranty that the buyer was arguably aware of prior to closing.

The recent ruling of the Delaware Court of Chancery in In re Dura Medic Holdings, Inc. has finally provided certainty in Delaware.^[1] However, cross-border M&A lawyers should know that the question remains open in Canada.

Sandbagging: Market Practice

The possibility of sandbagging presents M&A parties with three options: they can (1) include a “pro-sandbagging” clause that expressly permits it, (2) include an “anti-sandbagging” clause that expressly prohibits it, or (3) forgo any sandbagging clause and remain silent on the issue.

Which approach is most common? According to the most recent American Bar Association private target M&A deal point studies, the answer is clear: remaining silent is the most common avenue.^[2] In the United States, 76 percent of deals were silent on sandbagging, with 19 percent of deals including a pro-sandbagging clause and 5 percent of deals including an anti-sandbagging clause. In Canada, 82 percent of deals were silent on sandbagging, with 10 percent of deals including a pro-sandbagging clause and 8 percent of deals including an anti-sandbagging clause.

The question that follows is this: What approach to sandbagging would a court take when faced with contractual silence? The answer is now certain in Delaware. Things unfortunately remain murky in Canada.

Sandbagging in Delaware: Finally, Crystal Clear Skies

In Dura Medic Holdings, the buyer claimed damages for the seller’s breach of a representation and warranty stating that the target had not received notice of noncompliance with health-care laws in the preceding three years. The seller’s disclosure schedules had identified one such notice, but post-closing the buyer discovered others, and the additional regulatory review resulted in significant expense. The seller defended the claim on the basis that it had informed the buyer of the additional notices during a pre-closing due diligence call. The purchase agreement was silent on sandbagging. This put sandbagging squarely before the Delaware Court of Chancery.

Previous Delaware rulings had waffled somewhat. Historically, U.S. deal lawyers were confident that Delaware was a pro-sandbagging jurisdiction even without precedent directly on point. However, an aside by the Delaware Supreme Court in 2018 caused confusion by stating that “[v]enerable Delaware law casts doubt” on a buyer’s ability to sandbag.^[3] Subsequent and more favorable obiter comments by Delaware courts on sandbagging calmed U.S. lawyers’ concerns.^[4]

But it was not until Dura Medic Holdings, decided earlier this year, that the issue of sandbagging was confronted head-on. The court rejected the seller’s defense, explaining that the seller’s disclosure during the due diligence call “has no bearing on the legal analysis.”^[5] The reason was that a “breach of contract claim is not dependent on a showing of justifiable reliance.”^[6] The result was that having “contractually promised [the buyer] that it could rely on certain representations, [the seller] is in no position to contend that [the buyer] was unreasonable in relying on [the seller’s] own binding words.”^[7]

Notwithstanding that the purchase agreement was silent on sandbagging, the alleged knowledge of the buyer at execution that the representation was inaccurate did not impact the buyer’s ability to later claim for a breach of the representation. Stated more simply, Delaware is a pro-sandbagging jurisdiction.

Sandbagging in Canada: Still Cloudy, Twenty Years and Counting

The law around sandbagging in Canada is much less clear. The principal reason is conflicting appellate precedent from over two decades ago. Neither ruling was directly on point, but their general implications for sandbagging are relatively uncontroversial. The first, a 2001 ruling of the Alberta Court of Appeal (“ABCA”), appeared to endorse sandbagging.^[8] The second, a 2003 ruling of the Ontario Court of Appeal (“ONCA”), cast doubt on it.^[9]

The waters have since been muddied further by related rulings by the Supreme Court of Canada (“SCC”) regarding the duty of good faith in contract in 2014^[10] and 2020.^[11] The first established good faith as a “general organizing principle” of the common law, which includes a duty of honest performance in contract.^[12] It explained that this includes a duty that parties “not lie or otherwise knowingly mislead each other about matters directly linked to the performance of the contract.”^[13] The second elaborated that dishonest performance may include “lies, half-truths, omissions, and even silence, depending on the circumstances.”^[14]

Neither ruling addresses sandbagging. But it is arguable that the SCC’s treatment of good faith and honest performance could undermine an attempt at sandbagging, along the lines of the ONCA ruling. Alternatively, a court could instead focus on the SCC’s instruction that good faith does not require a counterparty to “forego advantages flowing from the contract,”^[15] along the lines of the ABCA ruling.

Practical Takeaways for Cross-Border Private M&A

Sandbagging raises complicated issues. Luckily for U.S. M&A lawyers, things have been greatly simplified by Dura Medic Holdings. On the other hand, given the continued uncertainty in Canada, U.S. M&A lawyers should remain vigilant regarding the potential complications surrounding sandbagging when working on cross-border deals. Given that the large majority of Canadian private M&A agreements go silent on sandbagging, these potential complications will quickly take on practical, real-world significance should a buyer foresee a possibility of sandbagging or should the seller suspect that the buyer may engage in sandbagging.

M&A parties should also be aware of related practical considerations relevant to sandbagging. For example, when acting for the buyer and the agreement is either silent on sandbagging or includes an anti-sandbagging clause, lawyers should be alert to the possibility that the seller might be tempted to engage in a “document dump” whereby a large amount of disclosure is made in the run-up to execution or closing. Related considerations regarding the drafting of the purchase agreement include (1) cross-referencing and whether disclosure in one disclosure schedule may constitute disclosure with respect to related disclosure schedules and (2) whether the acquisition agreement addresses updates to the seller’s disclosure schedules prior to closing (in the event of a transaction that has an interim period between signing and closing). Specific considerations regarding the latter include whether the agreement is silent on the point, expressly prohibits it, or expressly permits or requires it. This is important because whether (and under what conditions) the seller is permitted to update its disclosure schedules post-execution but prior to closing can impact the chance of sandbagging issues later arising.

1. In re Dura Medic Holdings, Inc., No. 2019-0474-JTL, 2025 Del. Ch. LEXIS 47 (Jan. 29, 2025). ↑

2. See the 2023 ABA Private Target Mergers & Acquisitions Deal Points Study (including transactions from 2022 and Q1 2023) and the 2025 ABA Canadian Private Target Mergers & Acquisitions Deal Points Study (including transactions from 2020, 2021, and 2022). ↑

3. Eagle Force Holdings, LLC v. Campbell, No. 10803, 2018 Del. LEXIS 233, at *66 (May 24, 2018). ↑

4. See, e.g., Arwood v. AW Site Servs., No. 2019-0904-JRS, 2022 Del. Ch. LEXIS 57 at *7 (Mar. 9, 2022) (“In my view, Delaware is, or should be, a pro-sandbagging jurisdiction.”). ↑

5. Dura Medic Holdings, 2025 Del. Ch. LEXIS 47, at *40. ↑

6. Id. at *41. ↑

7. Id. at *42. ↑

8. Eagle Res. Ltd. v. MacDonald, 2001 CanLII 264 (Can. Alta. C.A.). ↑

9. Transamerica Life Can. Inc. v. ING Can. Inc., 2003 CanLII 9923 (Can. Ont. C.A.). ↑

10. Bhasin v. Hrynew, 2014 CanLII 71 (S.C.C.), [2014] 3 S.C.R. 494. ↑

11. C.M. Callow Inc. v. Zollinger, 2020 CanLII 45 (S.C.C.), [2020] 3 S.C.R. 908. ↑

12. Bhasin, 2014 CanLII 71. ↑

13. Id. ↑

14. Callow, 2020 CanLII 45. ↑

15. Bhasin, 2014 CanLII 71. ↑

When an insured is pursuing a representation and warranty insurance (“RWI”) claim, a critical consideration is whether diminution in value damages (“DIV Damages”) can be asserted as Loss covered by the RWI policy.^[1] This article, to be published in four parts, discusses Delaware mergers and acquisitions (“M&A”) damages law regarding DIV Damages and describes how an insured can pursue them as part of an RWI claim.

Part I addresses (i) the principal differences between DIV Damages calculated using a multiple of EBITDA methodology (“MOE Methodology”) and DIV Damages calculated using a discounted cash flow methodology (“DCF Methodology”), and (ii) the evolution of cases involving DIV Damages calculated using an MOE Methodology under Delaware M&A damages law. Part II will address the evolution of cases involving DIV Damages calculated using a DCF Methodology under Delaware M&A damages law. Part III will discuss the requirements for a DIV Damages award as part of an RWI claim. Part IV will discuss the limitations on, and other matters regarding, a DIV Damages award as part of an RWI claim.

Each part of this article contains practice tips for attorneys for insureds seeking recovery of DIV Damages as part of an RWI claim.

What Are DIV Damages?

DIV Damages are a form of expectation damages in which the amount of the damages is the difference between (i) the value of the target business as represented to the buyer (almost always the purchase price paid for the target business by the buyer) and (ii) the value of the target business after giving effect to the diminution in the target business resulting from a breach of the Acquisition Agreement representations and warranties (“R&W Breach”) or from fraudulent misrepresentation or deceit regarding the target business. DIV Damages are often the most significant portion of an RWI claim recovery sought by an insured.

Notwithstanding the terminology often inaccurately used, DIV Damages are not “multiplier losses” or based on “multiplied damages.” Instead, DIV Damages are M&A damages calculated by reference to either: (i) in the case of an MOE Methodology, (a) an actual or deemed shortfall in the EBITDA of the target business for a specified measurement period (“Measurement Period EBITDA”) caused by the R&W Breach or by the fraudulent misrepresentation or deceit, times (b) the multiple applied by the insured to the Measurement Period EBITDA in determining the purchase price to pay for the target business; or (ii) in the case of a DCF Methodology, the loss of future cash flows and terminal value over a specified period caused by the R&W Breach or by the fraudulent misrepresentation or deceit, discounted to present value by the application of a discount factor. By comparison, an example of “multiplied damages” would be treble damages awarded to a private plaintiff for an antitrust violation, in which damages are calculated and then statutorily multiplied by three to determine the aggregate amount to be paid by the offending person or entity to the private plaintiff for the specified antitrust violation.^[2]

The Principal Differences Between Using an MOE Methodology and Using a DCF Methodology to Calculate DIV Damages

An MOE Methodology and a DCF Methodology are both ways to value a target business. However, on the one hand, an MOE methodology is retrospective, focused on historical EBITDA for a specified Measurement Period prior to the Acquisition, typically a specified calendar year or a trailing twelve months (“TTM”) or latest or last twelve months (“LTM”) period before a specified month-end date. On the other hand, a DCF methodology is prospective, focused on projected cash flows during a specified period—say five years or seven years after the Acquisition—with a terminal value (i.e., an additional number of years of projected cash flows summed) added on, all subject to a discount rate. As such, DIV Damages calculated using a DCF methodology are often confused with, and even referred to as, “lost profits damages.”

In the case of an MOE Methodology, it is the multiple that accounts for the risks associated with the target business. In the case of a DCF Methodology, it is either the projected cash flows themselves or the discount rate (sometimes referred to as a “risk-adjusted discount rate”) that accounts for the risks associated with the target business. In the case of an MOE Methodology, it is the multiple that increases the size of the shortfall in EBITDA during the Measurement Period from 1x. In the case of a DCF Methodology, it is the multiple years of projected cash flows and the terminal value that increase the size of the loss of cash flows from 1x.

One other note about the two methodologies: While this is not true across the board, financial buyers such as private equity firms tend to use an MOE Methodology to determine the purchase price for the target business (subject to deal adjustments, as in the case of an auction), while strategic buyers tend to use a DCF Methodology to confirm the acceptability of, if not actually establish, the purchase price for the target business.

Evolution of Cases Involving DIV Damages Calculated Using an MOE Methodology Under Delaware M&A Damages Law

From the Beginning: Cobalt v. Crystal

The seminal case involving DIV Damages calculated using an MOE Methodology under Delaware M&A damages law is the 2007 Delaware Chancery Court case of Cobalt v. Crystal.^[3] Cobalt involved the sale of WRMF—a West Palm Beach, Florida, radio station—by James Crystal Enterprises, LLC (“Crystal”) to Cobalt Operating, LLC (“Cobalt”), for a purchase price of $70 million. The purchase price was based on measurement period Broadcast Cash Flow (“BCF”) (a “financial measure . . . roughly akin to [EBITDA]”^[4]) of the target of $5 million, times a multiple of 14.

After finding that Crystal had breached and committed fraud with respect to certain Acquisition Agreement representations and warranties, then–Vice Chancellor Strine^[5] awarded Cobalt (i) DIV Damages of $11 million, based on (a) a measurement period BCF of $4.2 million adjusted to eliminate the effects of the R&W Breach, times (b) the multiple of 14 ($58.8 million, rounded up to $59 million, and then subtracted from the purchase price paid of $70 million); (ii) indemnification for out-of-pocket damages of $180,754 for credits extended by Cobalt to certain advertisers in respect of the R&W Breach; (iii) prejudgment interest; and (iv) reasonable attorney fees and costs.^[6]

In support of the award of DIV Damages to Cobalt, Vice Chancellor Strine made a number of factual and legal findings, including the following:

• “the traditional method of computing damages for a breach of contract claim is to determine the reasonable expectations of the parties”
• “[e]xpectation damages [in this case, the DIV Damages] are calculated as the amount of money that would put the non-breaching party in the same position that the party would have been in had the breach never occurred”
• Crystal “knew Cobalt was relying on a cash flow multiple in reaching the price it was willing to pay for WRMF”
• “Crystal did not present its own valuation evidence” to counter the valuation evidence presented by Cobalt^[7]

Cobalt continues to stand as the preeminent Delaware M&A damages law case involving an award of DIV Damages calculated using an MOE Methodology for an R&W Breach.

I Thought There Was a Rule Against Perpetuities? Zayo v. Latisys

The next significant Delaware M&A damages law case involving DIV Damages calculated using an MOE methodology is the 2018 Delaware Chancery Court case of Zayo v. Latisys.^[8] Zayo involved the sale of a group of IT infrastructure service providers ( “Latisys Companies”) by seller Latisys Holdings, LLC (“Latisys Holdings”) to buyer Zayo Group, LLC (“Zayo”) for a purchase price of $675 million. Zayo performed “extensive financial modeling of the Latisys business,” including “a synergies analysis[,] . . . an analysis of implied multiples of publicly known comparable transactions[,] . . . a 30-year [DCF] analysis, a net present value sensitivity analysis and an internal rate of return analysis.”^[9]

Vice Chancellor Slights first determined that the buyer Zayo had failed to prove a breach of the Acquisition Agreement representations and warranties, which by itself was dispositive of the case and did not require him to consider Zayo’s M&A contract damages claim. Nevertheless, “for the sake of completeness,”^[10] Vice Chancellor Slights also determined that Zayo was not entitled to a recovery of DIV Damages in respect of the asserted R&W Breach. In support of that determination, he made a number of factual and legal findings, including the following:

• Zayo’s valuation expert “lack[ed] . . . experience in valuing going concern businesses”
• Zayo’s valuation expert “base[d] her opinion solely on a multiple of EBITDA” analysis, even though “there [was] no evidence that Zayo actually based its purchase price on a multiple of EBITDA”
• to justify DIV Damages, “[t]he actual value the [buyer] received . . . must assume, and account for, a diminution of the [target’s] earnings into perpetuity”
• “[t]he ‘benefit of bargain’ methodology is appropriate for calculating damages only when the alleged breach of the representation or warranty has caused a permanent diminution in the value of the business (as a result of lost revenues into perpetuity) and the business has thereby been permanently impaired. This is where Zayo’s proof, and [its valuation expert’s] damages calculations, fell short.”^[11]

Although Vice Chancellor Slights’s damages dictum (that the loss and repricing of short-term customer contracts did not support Zayo’s claim for DIV Damages) was justified on the facts of the case, the extent and definitiveness of his findings that only a loss of revenues “into perpetuity” and thus a “permanent impairment” of earnings can support a claim for DIV Damages has been viewed by many M&A practitioners as legally unjustified.^[12] Nevertheless, for a period of time, RWI claimants seeking recovery of DIV Damages had a difficult precedent to contend with in Zayo, absent evidence of such a loss into perpetuity and permanent impairment.

The End of Perpetuity? In re Dura Medic

And then came the 2025 Delaware Chancery Court case of In re Dura Medic.^[13] Dura Medic involved the sale of a durable medical equipment supply company (“Dura Medic”) by its stockholders (“DM Stockholders”) to Comvest Partners, a private equity firm (“Comvest”), for a purchase price of $30 million. The buyer Comvest based the purchase price on the Company’s LTM^[14] April 2018 EBITDA, times a multiple of 6.7797.^[15]

After determining that certain Acquisition Agreement representations and warranties had been breached, Vice Chancellor Laster awarded Comvest: (i) DIV Damages of $2,847,890, based on (a) a deemed shortfall in Measurement Period EBITDA of $433,322 resulting from lost profits from two customers at the heart of one of the two R&W Breaches, times (b) the multiple of 6.7797, minus (c) $89,903 in actual post-Acquisition collections from such customers; and (ii) $100,000 of out-of-pocket damages for fees paid to a healthcare regulatory consulting firm with respect to the other of the two R&W Breaches.^[16]

In connection with awarding DIV Damages to Comvest, Vice Chancellor Laster made a number of factual and legal findings, including the following:

• Where the Acquisition Agreement is silent on the appropriate measure of damages, “the court must look to the common law [, under which] a party can recover reasonable expectation damages based on a multiple where the [purchase] price was established with a market approach using a multiple.”^[17]
• With respect to the DM Stockholders’ asserting Zayo as a precedent that only permitted the application of a multiple of a shortfall in EBITDA to losses permanently affecting the target business:
  • the buyer Zayo had provided no evidence in that case that it had based the purchase price for the target Latisys Companies on a multiple of EBITDA
  • the customer contracts at issue in that case all expired in one year or less after the Acquisition
  • the case’s “reference to a ‘permanent diminution . . . into perpetuity’ . . . [did] not translate into a test for future cases”
  • the Zayo court had offered no legal authority for the “permanent diminution” requirement for the application of a multiple^[18]
• Revenue from new customers added after the Acquisition would have been additive if Dura Medic could also have retained the lost customers at the heart of the applicable R&W Breach.^[19]
• Fraud is not required to apply a multiple in calculating expectation damages.^[20]

To summarize the foregoing, Dura Medic not only distinguished the damages findings in Zayo but also concluded that a recurring diminution in EBITDA after the Acquisition, not a “permanent diminution . . . into perpetuity” of EBITDA, was sufficient to support the award of DIV Damages based on a multiple.

But is Zayo actually dead? Cobalt, Zayo, and Dura Medic were all Delaware Chancery Court cases, decided by three different Vice Chancellors over an eighteen-year period. While the Delaware Supreme Court did affirm Vice Chancellor Strine’s decision in Cobalt, the Delaware Supreme Court has not directly or indirectly overruled Vice Chancellor Slights’s damages decision in Zayo. All that said, the strength and logic of the DIV Damages findings in Dura Medic, particularly those directly contradicting the findings in Zayo, taken together with the dicta status of the DIV Damages findings in Zayo, do appear to toll a death knell for the Zayo requirement of a “permanent diminution . . . into perpetuity.”

Practice Tips for Attorneys for Insureds

In the Acquisition Agreement drafting and negotiation phase, consider the following:

• If the definition of Damages or Losses in the Acquisition Agreement is incorporated into the definition of Loss in the policy, ensure that the definition in the Acquisition Agreement does not exclude “diminution in value,” “multiple of EBITDA,” “multiplier damages,” “lost profits,” or the like, and preferably try to have the definition explicitly include the first or second of those terms.
• If one-half (or some other portion) of the policy retention can be borne by the seller through an indemnification escrow provided for in the Acquisition Agreement, try to ensure that no type of Damages or Losses are disclaimed or waived in the Acquisition Agreement that are covered by the policy.

This article is the first in the RWI Practice Insights series by John T. Capetta.

1. This article focuses on buyer-side RWI policies and U.S. law (principally Delaware case law). Although there may be other methodologies to calculate DIV Damages, this article focuses on those calculated using either an MOE Methodology or a DCF Methodology as defined in the second paragraph. The period of time for which the historical EBITDA is measured in an MOE Methodology, or for which the projections used in a DCF Methodology are included, is referred to in this article as the “Measurement Period.”

   • As used in this article:
     • the term Loss has the definition set forth in the RWI policy;
     • the term Acquisition Agreement includes stock purchase agreements, merger agreements, asset purchase agreements, and other types of business combination agreements by which a buyer acquires a target business from a seller;
     • the term Acquisition refers to the business combination contemplated by the Acquisition Agreement;
     • the term the buyer and the term the insured are often used interchangeably;
     • the term target and the term target business are used interchangeably;
     • the term R&W Breach also includes a claim under an RWI policy with respect to a tax indemnification provision in the Acquisition Agreement; and
     • the phrase without required disclosure by the seller refers to a failure by the seller to make a disclosure to the buyer even though required to do so by a representation and warranty in the Acquisition Agreement.

   ↑

2. See Section 4 of the Clayton Antitrust Act of 1914 with respect to certain violations of the U.S. antitrust statutes. 15 U.S.C. § 15. ↑

3. Cobalt Operating, LLC v. James Crystal Enters., LLC, No. 714, 2007 WL 2142926 (Del. Ch. July 20, 2007), aff’d, 945 A.2d 594 (Del. 2008) (unpublished table decision). ↑

4. Id. at *1 n.1. ↑

5. Vice Chancellor Strine later became the Chief Justice of the Delaware Supreme Court. Vice Chancellor Strine issued some of the most significant M&A opinions ever, and certainly some of the most entertaining. See, e.g., his description and the depiction of the “Rose Mary Stretch” in Cobalt, 2007 WL 2142926, at *2. ↑

6. Cobalt, 2007 WL 2142926, at *4, *29–31. ↑

7. Id. at *29 (footnotes omitted). For want of a better place in this article to discuss it, the 2013 Delaware Chancery Court case of Universal Enterprise Group, L.P. v. Duncan Petroleum Corp., No. 4948, 2013 WL 3353743, at *19 (Del. Ch. July 1, 2013), aff’d, 99 A.3d 228 (Del. 2014) (unpublished table decision), contains potentially confusing language that describes diminution in value as an “alternative to expectation damages in particular contexts,” rather than as a type of expectation damages. However, in Universal, expectation damages would arguably have been the costs to restore certain parcels of real property (which made up the target business) to their expected condition as represented by the seller in the Acquisition Agreement, with diminution in value damages being an alternative method to compensate the buyer for the injury suffered as a result of the R&W Breaches in question. The Chancery Court rejected diminution in value damages in part because they would have produced a damages award “an order of magnitude greater than an award based on expectation damages” and thus would have been “disproportionate, constitute economic waste, and bestow a windfall.” Id. at *20. In the end, the Chancery Court awarded neither expectation damages nor diminution in value damages but instead only “actual damages” (to make matters even more confusing) in the form of the costs and expenses incurred by the buyer in remediating the parcels of real property after the Acquisition. ↑

8. Zayo Group, LLC v. Latisys Holdings, LLC, No. 12874, 2018 WL 6177174 (Del. Ch. Nov. 26, 2018). Then–Vice Chancellor Strine did issue another opinion after Cobalt and prior to Zayo that involved a multiple of EBITDA methodology: WaveDivision Holdings, LLC v. Millennium Digital Media Systems, L.L.C., No. 2993, 2010 WL 3706624 (Del. Ch. Sep. 17, 2010). However, WaveDivision arose from certain covenant breaches by the seller, which led to the seller’s not selling the target businesses to the buyer, rather than R&W Breaches. Moreover, WaveDivision involved a comparison of (i) the purchase price for the target businesses that the buyer would have paid had the Acquisition been consummated, to (ii) the value of the target businesses as they would have been operated by the buyer in the future, rather than their value ex ante (which arguably would have been equal to the proposed purchase price and thus have left the buyer without a real remedy). As a result, while there are aspects of WaveDivision that are instructive, such as with respect to mitigation, it does not fit into the line of Delaware M&A damages law cases regarding DIV Damages calculated using an MOE Methodology that began with Cobalt. ↑

9. Zayo, 2018 WL 6177174, at *6. ↑

10. Id. at *15. ↑

11. Id. at *15–17 (footnotes omitted). Vice Chancellor Slights had noted that the buyer Zayo’s “initial indication of interest [IOI] . . . ‘propose[d] a total value in the range of $625M–$655M in cash (approximately 11–11.5x Q4 2014E LQA [presumably “Last Quarter Annualized”] Adjusted EBITDA of $56.8M),’” id. at *4 (citing and quoting the IOI), but he inexplicably seemed to disregard that IOI in determining that “there is no evidence that Zayo actually based its purchase price on a multiple of EBITDA.” Id. at *17. ↑

12. See, e.g., E. Hutchinson Robbins, Jr., M&A Representation and Warranty Damages: The Myth of Lost Revenues into Perpetuity, Bus. L. Today (Aug. 19, 2021). ↑

13. In re Dura Medic Holdings, Inc. Consolidated Litigation, 333 A.3d 227 (Del. Ch. 2025). ↑

14. In Dura Medic, “LTM” stands for “Last Twelve Months.” Id. at 243. ↑

15. Id. at 259. ↑

16. Id. at 261–3. ↑

17. Id. at 259 (footnotes and internal quotation marks omitted). ↑

18. Id. at 259–60 (footnotes omitted), 260 n.55. In Zayo, Vice Chancellor Slights gave significant weight to the American Institute of Certified Public Accountants Mergers and Acquisitions Dispute Practice Aid’s pronouncements regarding the need for a permanent impairment in value in support of his determination that a multiple of EBITDA was not an appropriate methodology to calculate damages absent such a permanent impairment, but he did not cite any cases or other basis in the law for that determination, although he did distinguish a number of cases, including Cobalt, in reaching that determination. Zayo, 2018 WL 6177174, at *17 n.215. ↑

19. Dura Medic, 333 A.3d at 260. ↑

20. Id. ↑

Imagine a fintech startup that deploys an artificial intelligence (“AI”) model to underwrite small-business loans—only to face a demand letter alleging “intentional discrimination” under Texas law. On June 22, 2025, Governor Greg Abbott signed the Texas Responsible Artificial Intelligence Governance Act (“TRAIGA”), placing Texas at the vanguard of state-level AI regulation. As the fourth state to enact a comprehensive AI statute—after Colorado, Utah, and California—Texas now offers both a clear roadmap for developers and heightened risks for those whose AI decisions cause harm.

With the passage of TRAIGA, business trial lawyers are entering an era where AI regulation is no longer a theoretical concern but a live, litigable issue. For trial attorneys handling commercial disputes, TRAIGA’s unique blend of intent-based liability and centralized enforcement reshapes the evidentiary landscape, requiring more rigorous documentation and strategic foresight. Business lawyers must now anticipate how a client’s AI-related decisions prone to allegations of bias—especially in areas such as lending, hiring, and marketing—might be scrutinized under a standard that demands proof of knowing and intentional misuse.

TRAIGA, which takes effect January 1, 2026, introduces a suite of targeted prohibitions, regulatory mechanisms, and compliance frameworks that will shape not only the development and deployment of AI systems in Texas but also the broader landscape of business litigation and regulatory strategy. For business trial lawyers, understanding the contours of this new law is essential, as it will influence litigation strategy, evidentiary standards, and the future of AI-driven business operations.

The version of TRAIGA that was passed and signed into law represents a pared-down evolution from earlier, more expansive drafts that mirrored the risk-based frameworks of the EU AI Act and the Colorado AI Act. The final version, however, reflects a pragmatic shift toward outcome-focused regulation, emphasizing specific prohibited uses of AI while scaling back broad compliance mandates for the private sector. This approach has direct and nuanced implications for business litigation in Texas and potentially beyond.

Intent-Based Evidentiary Standard for Liability

Under Section 4 of TRIAGA, civil liability attaches only where a developer or deployer “intentionally” uses AI to

1. promote self-harm or suicide;
2. promote harming another person;
3. facilitate criminal activity;
4. engage in unlawful discrimination;
5. produce unlawful deepfakes or child-exploitation content; or
6. infringe, restrict, or impair constitutional rights.

The elevated burden of proof placed on plaintiffs is one of TRAIGA’s most consequential implications for business trial attorneys. This intent-based standard departs from risk-only frameworks (e.g., the EU AI Act), requiring claimants to adduce evidence of purposeful misconduct—not merely disparate outcomes. Traditional arguments relying on disparate outcomes or statistical disparities will no longer suffice without evidence of deliberate intent; claimants must now demonstrate that the developer or deployer acted to discriminate or cause harm intentionally.

This heightened standard may reduce the volume of AI-related business litigation premised on algorithmic bias, particularly in employment, lending, and other regulated sectors. It also makes early discovery strategy and internal documentation absolutely pivotal, placing a premium on robust documentation and internal controls. TRAIGA’s safe harbor provisions—such as affirmative defenses for parties that discover violations through internal review, adversarial testing, or compliance with recognized risk management frameworks (e.g., the National Institute of Standards and Technology (“NIST”) AI Risk Management Framework)—may encourage businesses to adopt proactive compliance measures and self-audit protocols. Attorneys representing businesses should proactively advise clients to develop and maintain records showing compliance with these recognized frameworks to activate the safe harbor protections and blunt any allegation of purposeful harm. Further, businesses should strengthen internal controls and preserve audit trails demonstrating their AI systems’ legitimate aims and operational safeguards. These records can become powerful tools during motions to dismiss or summary judgment.

TRAIGA vests exclusive enforcement authority in the Texas attorney general, precluding private rights of action. The attorney general is empowered to investigate alleged violations, issue civil investigative demands, and seek injunctive relief and civil penalties ranging from $10,000 to $200,000 per violation, with additional penalties for continuing violations. TRAIGA also provides for a sixty-day cure period following notice of a violation, incentivizing prompt remediation. This centralized enforcement model may streamline the adjudication of AI-related disputes, reduce the risk of inconsistent outcomes, and provide greater predictability for businesses.

Impact on Discovery and Pretrial Practice

TRAIGA’s focus on intent and its explicit exclusion of liability for unintended third-party misuse of AI systems may limit the scope of discovery into downstream uses. In traditional business litigation, especially in cases involving products or technologies with broad downstream applications, plaintiffs often seek extensive discovery into how a product was used by third parties, the range of possible outcomes, and the foreseeability of misuse. Such discovery can be costly, time-consuming, and burdensome, as it may require the production and analysis of voluminous data regarding end-user behavior, system outputs in varied contexts, and communications with customers or partners.

AI tools like Technology-Assisted Review (“TAR”) can bring order to an overwhelming amount of complex data that in recent years had complicated discovery and pretrial motion practice while creating a massive litigation expense burden. TRAIGA clarified the lines as to what behavior is culpable, which reduces disputes over the adequacy of internal controls and the potentially limitless downstream effects of AI systems in the hands of third parties. And because TRAIGA excluded TAR from its scope, it implicitly affirms continued use of TAR to streamline discovery of just the relevant evidence.

Potential Benefits and Drawbacks of TRAIGA’s Sandbox Regulatory Model

One of TRAIGA’s most innovative features is the establishment of a regulatory sandbox program, administered by the Texas Department of Information Resources in consultation with the Texas Artificial Intelligence Council established by the legislation. The sandbox allows approved participants to develop and test AI systems in a controlled environment, temporarily exempt from certain state licensing and regulatory requirements, for up to thirty-six months.

The sandbox model offers regulatory flexibility and innovation by providing a structured pathway for businesses to experiment with novel AI applications without the immediate risk of regulatory penalties, which is particularly valuable in areas of legal uncertainty where traditional regulatory frameworks may lag behind technological advances. Additionally, by requiring quarterly reports on system performance, risk mitigation, and stakeholder feedback, the sandbox generates empirical data that can inform future legislative and regulatory reforms, enhancing the capacity of regulators and lawmakers to craft targeted, effective policies. Moreover, the Texas sandbox positions the state as a leader in AI regulation, potentially serving as a model for other jurisdictions and facilitating cross-jurisdictional data analysis, which may lay the groundwork for future reciprocity agreements, enabling businesses to scale innovative AI solutions across state lines with greater legal certainty. For business law practitioners, the sandbox offers a clear, time-limited framework for managing regulatory risk during the development and deployment of cutting-edge AI systems, and participation in the sandbox may also serve as a mitigating factor in enforcement actions, further incentivizing compliance.

However, the sandbox model also presents potential drawbacks. While it promotes innovation, it might contribute to a fragmented regulatory landscape if other states adopt divergent models or standards, creating challenges to harmonizing compliance across jurisdictions for businesses operating nationally. Effective oversight of the sandbox requires significant administrative resources, including technical expertise and ongoing monitoring, potentially creating barriers to entry for smaller businesses and limiting the program’s inclusivity. While the future of state-level AI sandboxes is safe for the moment, it remains uncertain due to the prospect of federal preemption. Though one had been proposed, there was no state AI law moratorium in the recently signed federal budget bill; it was stripped out of the U.S. House version by the U.S. Senate. Still, the political forces that had it included in the first place may try again. This could suspend or nullify the Texas sandbox and similar programs, making it essential for business law professionals to closely monitor federal developments, as the regulatory environment may shift.

Practical Advice for Business Law Practitioners

Ultimately, TRAIGA isn’t just a compliance statute—it’s a blueprint for how AI liability will be litigated. For business trial lawyers, this signals a shift in how risk is assessed, evidence is preserved, and cases are pled. Those who understand TRAIGA’s enforcement structure, sandbox incentives, and documentation expectations will not only better defend their clients but also shape emerging jurisprudence on AI accountability. In this evolving landscape, the ability to speak the language of both law and machine will become a key differentiator in the courtroom.

TRAIGA marks a watershed moment in the evolution of AI regulation. By focusing on specific, outcome-based prohibitions and embracing innovative regulatory mechanisms such as the sandbox model, Texas has crafted a framework that attempts to balance the imperatives of innovation, consumer protection, and legal certainty.

For business law practitioners, TRAIGA’s intent-based liability standard and robust safe harbor provisions offer both challenges and opportunities. Practitioners should consider advising clients to do the following:

• Maintain clear records of the intended uses and operational controls of AI systems to support defenses against claims of intentional misconduct.
• Implement and document compliance with frameworks such as the NIST AI Risk Management Framework to avail themselves of statutory safe harbors.
• Stay abreast of federal and state legislative activity, particularly regarding potential preemption and the evolution of sandbox programs.
• Evaluate whether to join the regulatory sandbox for innovative AI projects, balancing the opportunities for experimentation against the administrative requirements and potential for regulatory change.

This is the seventh installment in the Year in Governance Series from the In-House Subcommittee of the ABA Business Law Section’s Corporate Governance Committee. Each month, the series will share key tips on a different corporate governance topic. To get involved in the Corporate Governance Committee, please visit the committee’s webpage.

A message from Kathy Jaffari: “As Chair of the Corporate Governance Committee, I would like to extend my sincere appreciation to the authors for this publication. The Corporate Governance Committee has ongoing opportunities for writing and volunteering with various projects, whether it’s an article you want to publish or a CLE that you want to present. Our Committee is dedicated to helping you promote informative resources for corporate governance practitioners. You may contact me at [email protected] to get involved.”

The form and substance of board materials, aligned to a well-developed board agenda, are critical in enabling board members to effectively oversee the management of the company’s business, meaningfully engage in strategy and risk discussions, and satisfy their fiduciary duties, especially the duty of care. Well-organized and timely provided board materials are essential to ensure that directors are knowledgeable, prepared, and focused on the most significant issues, and to evidence the satisfaction of applicable corporate governance requirements.

1. Focus on the objectives. Board decks and memoranda, particularly those that discuss key strategies, should be aligned to the overall objectives for the particular board session or committee meeting for which they are prepared. Consideration should be given as to whether the matter being presented is for board approval, for discussion, or for informational purposes, and materials should clearly indicate why they are being included.
2. Implement a consistent format. A uniform, consistent format should be used for all board materials, including using clear and concise language, executive summaries, tables of contents, headings, graphics and visual aids, bold and/or underlined type to highlight key information, and appendices or glossaries. Utilizing a consistent format will enable directors to more easily navigate board materials and prioritize important information in preparation for key decision points and discussions.
3. Include necessary information while avoiding overload. Board materials should include the key information required to inform and prepare board members, including financial data and other important information and metrics regarding key strategies and risks. However, because overloading directors with too much information can be as counterproductive as providing too little information, board materials should not include excessive or irrelevant information or repetitive data that would be better presented in a summarized format or as a read-only item.
4. Provide context and use plain English. Because board members are not involved in the day-to-day management of the company, it can be helpful to provide context on the topic presented or a reminder that ties back to a discussion from a prior board meeting. Consider using one-slide summaries or executive summaries to provide directors with basic background on the topic presented and how it connects to the overall objectives of the meeting and, if applicable, to prior board discussions. In addition, avoid using industry jargon or acronyms without explanation. When dealing with complex topics, consider using a glossary that defines key terms, phrases, and concepts.
5. Confirm accuracy—especially if using AI. Management should take steps to confirm that board materials contain accurate information and present a complete picture, including information that is both positive and negative for the company. With the increased use of artificial intelligence (“AI”), it is even more important to confirm accuracy, given that some AI platforms are known to generate inaccurate, incomplete, or out-of-date content.
6. Provide adequate time for review. Board materials should be circulated pursuant to an established practice that provides adequate lead time for directors to carefully review the materials, form opinions prior to board meetings, and raise questions that may be answered or discussed in advance of meetings. Providing complex materials relating to financial or technical topics without adequate notice could give rise to allegations that decisions were made without full understanding or consideration of the relevant factors. While companies may take different approaches, distributing board materials at least one week in advance of board meetings would be consistent with applicable best practices.
7. Distribute materials securely. If possible, leverage a secure board management platform for distributing board materials in advance of meetings. Many vendors offer board management solutions that offer cybersecurity controls, access controls, and integration with company record storage and collaboration platforms. Additional best practices include providing instructions that directors may follow to access board materials and confirming with directors that they have been able to access board materials after they have been posted and made available.
8. Consider access and searchability. Another advantage of using a board management platform is that most vendors offer a way for directors to quickly access board materials from past meetings and search past board materials for specific documents or issues. This saves directors time when preparing for board meetings, as many board and committee topics will constitute regular agenda items or updates from a prior board meeting.
9. Destroy drafts. Drafts of board decks and memoranda can potentially be discoverable in subsequent litigation focusing on the actions or decisions of the board. Thus, it’s best to put in place a process for the destruction of draft decks, memoranda, and other materials other than the final versions that were shared with the board or a board committee. If drafts are available and are subsequently produced, changes to the content of the materials can be taken out of context and misconstrued.
10. Seek feedback and adjust. Following board meetings, management should seek feedback from directors (and should share feedback internally) regarding the form and substance of board materials, including which materials were effective and where there could be opportunities for improvement going forward. Feedback on board materials can also be requested through the annual board evaluation process. As new directors join the board and new issues are presented at meetings, ensuring that board materials hit the mark will always be an iterative process.

The views expressed in this article are solely those of the authors and not their respective employers, firms, or clients.

The corporate alternative minimum tax (“CAMT”) requires noncorporate taxpayers—e.g., partnerships, limited liability companies, and S corporations (“noncorporate entities”)—to compile unique accounting books (“CAMT books”) for high-income corporate interest holders (“CAMT corporations”) for CAMT purposes. This process is time-intensive and expensive, and it requires backdating to years before the CAMT corporation obtained an ownership interest in the noncorporate entity. This requirement may even be triggered regardless of how many other entities are in the ownership chain between the noncorporate entity and eventual CAMT corporation.

Under model governance documents, the cost burden of this process is on the noncorporate entity. Drafters of governance documents should consider the potential tax liability of all interest holders and potential interest holders when drafting governance documents. Failure to consider CAMT issues when preparing governance documents could cost founders in the long run.

CAMT: A Primer

CAMT came into effect under the Inflation Reduction Act of 2022. CAMT, generally speaking, is a 15 percent minimum tax on large corporations. CAMT applies to corporations making an average of $1 billion a year over a three-year span or $100 million a year over a three-year span for certain U.S. subsidiaries of foreign-parented groups (i.e., CAMT corporations). This amount is calculated using adjusted financial statement income (“AFSI”)—that is, accounting book income with certain adjustments. AFSI requires a third set of accounting books (i.e., CAMT books) to be prepared and maintained (in addition to regular accounting books and tax books). Thus, if a corporation, based on its AFSI, makes $1 billion a year, then the corporation must pay at least $150 million in federal income tax and compile CAMT books. If the standard federal corporate tax system does not require a payment of at least $150 million, CAMT requires a true-up tax bringing the total amount paid to $150 million.

At first glance, the CAMT framework appears narrow and irrelevant to most businesses, but there are a few crucial considerations that drastically expand CAMT’s impact. First, the $1 billion and $100 million thresholds are not inflation indexed, so the scope will expand over time. Second, the CAMT books must be maintained for any entity that may eventually be captured by CAMT because the preparation of the books would otherwise require backdating. Third, and most expansive, in order for a large corporation to fully prepare its own CAMT books, it must receive accurate income information from its pass-through subsidiaries prepared as CAMT books. This information must be provided even if the pass-through subsidiary itself is not subject to CAMT. Further, the requirement for the pass-through subsidiaries to prepare the books applies all the way up the chain to the CAMT corporation. In other words, every pass-through subsidiary controlled by a CAMT corporation (no matter how small the subsidiary, how attenuated the relationship, or whether the CAMT corporation is a minority owner) must prepare its own CAMT books.

Noncorporate Entity Burden

Generally, noncorporate entity operating agreements require the noncorporate entity to bear the cost burden of preparing tax information for its owners. For example, section 8.4 of the LexisNexis form operating agreement provides thus: “The Company shall send or cause to be sent to each Member within ninety (90) days after the end of each taxable year (i) such information as is necessary to complete the Member’[s] federal and state income tax returns.”^[1]

There are two approaches that the noncorporate entity could take in response to standard tax information operating agreement language: (1) bear the cost of preparing CAMT books for the CAMT corporation or (2) give all documents necessary for preparing CAMT books to the CAMT corporation, no matter how minor or attenuated the CAMT corporation’s interest.

First, if the noncorporate entity bears the cost burden of preparing CAMT books, this disproportionately harms the interest holders who do not have to comply with CAMT. These noncorporate entity interest holders must share, through their ownership interest, in the cost of CAMT compliance without receiving any benefit in return.

On the surface, this looks like an issue only for larger entities with corporate parent groups, but failing to consider CAMT expenses at the formation stage could cost founders in the long run. Further, there are many ways that a CAMT corporation could eventually become involved in a closely held business: (i) the founders could sell a portion of their ownership interest, (ii) a bankruptcy proceeding could result in the forced sale of an economic interest, or (iii) a founder could transfer their economic interest. Economic interests may generally be freely transferred. This can easily, sometimes over the objection of other owners, give a CAMT corporation (or any subsidiary of the CAMT corporation, no matter how distant) a taxable interest in the closely held business.

Second, under most standard governance documents, the noncorporate entity’s obligation can be satisfied by providing the CAMT corporation with sufficient information to complete the CAMT corporation’s federal tax returns. In most cases, however, simply giving the CAMT corporation enough information to compile the CAMT books itself is not an acceptable solution because that would transmit all, or nearly all, of the noncorporate entity’s financial documents to the CAMT corporation, even if the CAMT corporation is a minority owner.

Other Alternative Minimum Taxes

This tax information issue is not unique to CAMT: Organisation for Economic Co-operation and Development (“OECD”) Pillar II also requires the creation of separate accounting books and imposes a minimum tax on certain businesses. In the case of OECD Pillar II, those businesses are large multinational businesses. The same CAMT tax information and compliance cost issues arise for OECD Pillar II—that is, when a small company that does not need to comply with OECD Pillar II is owned in some part by a company that must comply with OECD Pillar II, the small company faces issues similar to being owned by a CAMT corporation.

CAMT followed in the footsteps of OECD Pillar II; these alternative minimum taxes are becoming more popular as the potential to exploit traditional corporate tax structures becomes more common. With this popularity, the issue of preparing entirely separate books for the different accounting structures of all entities in an ownership structure will only continue to proliferate. Further, there is nothing to prevent these alternative minimum tax structures from applying to smaller companies in the future.

Risk Mitigation for Alternative Entities

Going forward, governance documents should carve out CAMT-related expenses from a noncorporate entity’s obligation to provide tax information. If a company only engages with certain tax regimes (e.g., CAMT or OECD Pillar II) because of the status of one or more of its members, the cost of engaging with those tax regimes should be on those members. If the noncorporate entity, as a whole, bears the cost of engaging with those tax regimes, then this cost unrightfully diminishes the earnings of the interest holders of the noncorporate entity who have no obligation to engage with CAMT or other similar tax regimes.

Furthermore, those responsible for the maintenance of governance documents should consider amendments that address CAMT and other similar tax regimes.

1. Operating Agreement (Member-Managed, Multiple Members) (DE LLC), LexisNexis, (emphasis added). ↑