The purpose of this article is to address the intersection between corporate governance and artificial intelligence (AI), and to explain why corporate board members must be familiar with basic AI concepts in order to fulfill their fiduciary duties. AI is permeating all industries, and directors must be aware of risks associated with the ever-evolving landscape of such technology. Thus, at least a basic understanding of AI and its applicability to a board’s particular industry is necessary for board directors to comply with their fiduciary duties and for effective corporate governance.
What Is AI?
AI is the capacity of a computer or electronic device to use characteristics commonly associated with human intelligence, including reasoning and learning from prior experiences. AI devices can be based upon algorithms consisting of rules for initial assessment and directions of next steps to take depending on the initial assessment, machine learning where devices learn autonomously based on initial rules and data, and even deep learning, which is learning through “the development of large artificial neural networks.” See https://www.zendesk.com/blog/machine-learning-and-deep-learning/. Based on the algorithm or learning method, the AI devices are able to process significant amounts of data and reveal patterns to solve problems, make decisions, perform complex maneuvers, or provide additional data for human consideration. Examples of AI in daily life include:
- Netflix and Pandora recommending movies and songs based on past watching and listening behavior;
- Amazon recommending products for purchase based on past shopping behavior and shopping behavior of other individuals who have purchased the same product;
- self-driving cars that are able to detect motion around the vehicle and maneuver the car safely in response; and
- the ever-present Siri and Alexa that continually learn and improve based upon natural-language inquiries and requests.
In addition to the above-noted examples, other AI examples include programs to analyze credit eligibility, decipher which e-mails may be spam, and detect fraud in banking and credit transactions.
Risks of AI
Along with the increasing presence of AI, there is an underlying fear many have, no doubt due in part to movies like “The Terminator,” and “I, Robot,” that one day human usefulness will be eliminated, and mankind will be taken over by AI machines. Although some jobs have already become obsolete (e.g., automated phone systems, with voice-recognition software replacing some customer-service operators), other risks associated with AI relate to data privacy and making some businesses obsolete.
In addition, as computers and machines continually evolve into deep learning, it may be more problematic to discern how massive amounts of unreliable data are interpreted through AI. Further, it seems that almost every day there is a new headline about companies being hacked; thus, there are increasing concerns about privacy and protection of data in an AI world.
Brief Background on Fiduciary Duties
Delaware law, which is viewed by many as the “nation’s corporate law,” requires members of boards of directors to comply with the fiduciary duties of loyalty and care. In Re PLX Tech. Stockholders Litig. 2018 Del. Ch. LEXIS 336, *64 (Oct. 16, 2018). That is, they have a duty to act in the best interests of the corporation and its shareholders, Frederick Hsu Living Trust v. ODN Holding Corp., 2017 Del. Ch. LEXIS 67, *43–*44 (Apr. 14, 2017), and they must be fully informed before making decisions on behalf of the company. McMullin v. Beran, 765 A.2d 910, 921 (Del. 2000).
There are also subsidiary duties that derive from the duties of loyalty and care. One subsidiary duty is that board members have the duty of oversight, which is sometimes referred to as a Caremark duty, after the name of the case that articulated it. In essence, the duty of oversight requires that directors have in place an effective reporting or monitoring system and an information system that allows them to detect potential risks to the company. Reiter v. Fairbank, 2016 Del. Ch. LEXIS 158, *18–*22. This information and reporting system has been described by the courts as part of a risk-management program that allows the directors to be properly informed and to become aware of any developing trends or activities at their company that may create a risk of liability for the company. Id.
Another subsidiary duty is the duty of nondelegation. Directors cannot delegate to nondirectors “in a very substantial way their duty to use their own best judgment on management matters.” Canal Capital Corp. v. French, 1992 Del. Ch. LEXIS 133, *5 (“It is settled law that directors may not delegate to others those duties that are ‘at the heart’ of the management of the corporation”).
Intersection of Fiduciary Duties and Consideration of AI
In order to comply with their duties of loyalty and care, boards should consider whether AI can be useful to improve the business practices of the organization, or whether it should be incorporated to assist the board in making informed decisions.
In the first instance, if AI will reduce employee workload or overhead costs in general by automating a transaction, it may have the potential to significantly increase profits for the business. AI is available to analyze trends in the marketplace to assist in marketing the business and targeting new consumers. If the board does not keep abreast of such technological advances, then it is not being materially informed or investigating developing trends to effectively govern the corporation, and thus failing to fulfill its duty of care. It is essential for board members to continually keep informed about developments of AI that may have transformative effects on the business or make a particular business less necessary. For example, use of ubiquitous, free GPS that adapts to wrong turns and online, live traffic mapping displaced the necessity for printed maps. Board members of companies that printed maps should have seen this development around the bend.
If a board considers incorporating AI, key factors include whether the existing data maintained by the company is suitable for the data mining that would be needed for the particular AI technology. Data management will be key for implementation of AI. In order to comply with the duty of care and oversight, the board must have an understanding of how data is obtained and maintained to understand whether AI will be effective, or whether overarching changes in data management and storage must be undertaken. Further, how to secure the data and protect it from hackers will be crucial in incorporating AI, as will audits about whether AI is accurately interpreting data.
Utilizing AI to enhance the board’s decision-making capabilities and analysis of data may soon be more commonplace. AI can be used to track overall trends in how the business is spending and allocating funds as well as to mine information and data for alignment with the overarching goals of the business in various departments.
In 2014, a venture capital firm claimed to have “appointed” an AI program called “Vital” to its board of directors. Vital sifted through research of drugs used to test age-related diseases and would then advise as to whether the firm should invest in the drugs. Although Vital was not a voting member, and although all boards can consider experts or various sources of information to assist in decision-making, appointing an AI program to the board was indicative of the role AI can play in the governance of a corporation. It is important to note, however, that in Delaware board members must be “natural persons,” so “appointing an AI program” as a board member for a Delaware corporation would be impermissible. DGCL § 141(b).
To the extent a board moves forward to adopt AI, it is crucial that the board does not delegate its essential management functions and rely solely upon AI in making decisions for the corporation. Doing so would be a prohibited delegation of its duties. Further, although many AI technologies can reach a decision based on their interpretation of data, keeping a record of how they reached that decision can be more problematic. AI should not be the sole source upon which board members rely in governing a business or making decisions.
Conclusion
This article has only scratched the surface of the extent of AI’s potential impact on fiduciary duties if a board fails to consider AI. As with any development that may create risk or bring value to its shareholders, every board, regardless of industry, should consider what impact AI may have on its business and its corporate governance.