Directors will be forgiven for feeling like AI is “everything everywhere all at once”. Yes, our reference to the 2022 Oscar winning film is intentional. Like the movie, AI defies easy categorization. It can be equal parts exhilarating and hallucinatory, impressive and uncomfortable.
Amid waves of great change, it can help to take a step back. To focus less on the granular and to take time to absorb the bigger picture. To consider not just where we might be in six months or a year, but where we might be in four or five years’ time.
To do so as relates to AI in the boardroom, we’ve been reading the latest forward-looking business and legal analysis across such well-respected publications as the Harvard Business Review, the Harvard Journal of Law & Technology, and the Canadian Business Law Journal (collectively, the Articles).[1] While each of the Articles merits a full read, we know that time is a scarce resource, and so we’ve summarized parts of them in this legal update.[2]
We do not do so in any attempt to prognosticate,[3] but only to note a common theme recurrent in the Articles: a new and increasingly complex human-technology dynamic, and what that might mean for senior executives and corporate governance as AI continues to permeate the business world. We conclude with some of our own observations, including where we see reason to question or push back on the Articles (see Concluding Comments).
Some highlights of the Articles’ forecasts include:
- AI becoming more embedded in executive roles themselves while overall executive focus increasingly shifts to optimizing human-AI integration.
- AI becoming central not just to board decision-making but also to the board’s oversight of the company’s human-AI architecture and AI culture.
- The board’s duty of care evolving to focus less on individual AI-enabled decisions and more on the overall design of the systems through which AI-enabled decisions are made.
This is the fourth episode in our “AI in the Boardroom” series. For our first three episodes, see here and here and here. For more Fasken capital markets thought leadership, visit our Capital Markets and M&A insights hub and subscribe.
The Future of AI in the Boardroom from Three Angles
The Articles look to the future of AI in the boardroom from three angles: (1) executive leadership, (2) boards and directors, and (3) corporate law and governance.
1. Executive Leadership
While much debate has focused on AI’s impact on entry-level and lower-level jobs, the Harvard Business Review article argues that significant change at the top of the organizational chart is also underway – both at the level of individual executive roles and in the overall management of the company by executive leadership.
As evidence, the job descriptions of more than 5,000 open executive roles from 2019 to 2025 were analyzed to identify trends in the competencies applicants were required to have. The “clear pattern” detected is that technology has been a “primary force reshaping the composition of the C-suite.” It’s not that AI is replacing executives, but that AI is becoming “embedded in their roles.” The examples provided include:
- CEOs “spend far more time on technology and talent than a decade ago.”
- COOs are expected to “understand automation and analytics, as much as logistics or operations.”
- CFOs must “interpret data science and scenario modelling rather than just accounting.”
- CHROs are moving from “managing people” to understanding “talent as a system”.
The article argues that, for corporate leadership as a whole, the question is not simply whether to deploy AI but how to redesign organizations, hierarchies and workflows to best harness AI’s benefits. This is seen as requiring focus on “optimizing the interface between people, data and machines”. Put differently, it’s about “orchestrating” the integration of human and AI “at scale.”
The article further argues that, in this environment, the leadership traits that will grow more important will not just be those skills that are harder to automate, but also those that are most valuable in “coordinating humans with machines”, including self-awareness, curiosity, empathy and integrity. A corollary of this for senior executives and their organizations is that culture will become an increasingly important source of competitive advantage. In other words, the article argues that firms should embrace AI with corporate culture built around those qualities that enable AI-nimbleness, such as “transparency, adaptability and speed.” Executives, in turn, may “spend less time approving and more time interpreting, coaching, and challenging both humans and algorithms.”
2. Boards and Directors
The Articles’ see similar implications for directors and their decision-making. A “shift from oversight to augmented decision-making systems” is envisioned as AI tools are integrated into boards’ analytical processes. Deliberations become more “multi-intelligence” as human experience is blended with AI compute.
The Articles ask if the very nature of the board could be redefined – could AI agents advance to such a level of sophistication and capability that they could serve as credible board members themselves?
While this notion is fantastical, one of the Articles makes the point that AI agents “sitting” on boards is already a present-day reality in “smaller, younger, AI-native companies.” Also, in a development that post-dates the Articles, Delaware has begun work on legislation that would create a new form of business entity that would be managed by AI agents rather than humans.
Arguably less fantastical is the Articles’ suggestion that oversight of the strategic direction of a company by its board will increasingly involve oversight over AI system design. It’s argued that this would follow from the Article’s forecast summarized above (see Executive Leadership above) that management by executives will become more about “orchestrating” the integration of human and AI “at scale.” Oversight by the board would therefore involve more oversight over AI culture as a competitive advantage and management’s plans for configuring “environments in which humans and intelligent machines outperform either alone.”
According to the Articles, AI should also assist towards a board’s more conventional oversight of corporate culture, performance and strategic direction. It should mitigate a board’s dependency on management for most of its information, “facilitating better questions and engagement.” It should also enable boards to “collect and synthesize data on corporate activities in real-time”. All of this should in turn assist directors in meeting their duty of care to the corporation (see Corporate Law and Governance below).
The Articles make two warnings for directors regarding the integration of AI into their oversight. These come from different directions. First, a key risk for boards is not replacement by AI, but over-reliance on it; directors must not over-delegate or outsource judgment to systems they do not sufficiently understand (see Corporate Law and Governance below). Second, boards should be careful not to fall behind management in terms of AI literacy. Should this occur as management teams become more AI-enabled and business operations become more AI-oriented, boards that don’t maintain sufficient AI-fluency “risk becoming increasingly disconnected from how value is actually created.”
3. Corporate Law and Governance
What might all of this mean for corporate law and governance?
It’s not controversial among the Articles that a board’s duty of care to the corporation as relates to AI requires continuous monitoring and adaptation. This starts with understanding the company’s particular AI risk exposure, both externally to third parties (e.g., consumer claims) as well as internally (e.g., potential self-inflicted cybersecurity events, as we discussed in Episode 3). Once understood, appropriate AI protocols must be put in place (as we discussed in Episode 2). These must then be adapted accordingly and proactively as developments warrant, e.g., as news of novel and adverse business-related AI incidents emerge and as AI-related legal issues are addressed by the courts.
The Articles caution, however, that the interaction between AI and the law may become increasingly complex.
A simple example of an outstanding question is the effectiveness of contractual waivers and similar risk mitigation mechanisms as warnings and disclaimers. Will developers and deployers of AI models be able to push down AI-related risks to consumer users? A much more complicated example is the disconnect highlighted by the Articles in the interaction of, on the one hand, those principles of corporate and commercial law dependent on human decision-making, knowledge and/or intent, and, on the other hand, AI systems that are increasingly autonomous and opaque. To date, courts have generally had little tolerance for defences based around a plea that “the AI did it”. The company remains responsible because it put the AI in place. That said, greater tension between legal concepts based on human decision-making and autonomous AI agents are anticipated going forward, and so legislative solutions to address gaps as they appear are foreseeable, which could go so far as AI-related strict liability.
A similar evolution is anticipated by the Articles around the scope and substance of the board’s duty of care.
As we discussed in Episode 1, AI is a double-edged sword for boards. The Articles agree: the right balance must be found. On the one hand, boards should use AI cautiously. They must ensure they’re using sufficiently capable AI tools and that such tools are being used within the AI’s competence and limitations. But boards must also be sure to continue to apply human judgment, scrutiny and critical thinking skills. Not doing so will neither satisfy their duty of care nor attract the protection of the business judgement rule: directors can’t effectively or substantially outsource their decision-making to AI. On the other hand, given AI’s undeniable utility, under reliance on AI could raise similar risk from the opposite direction: boards shouldn’t unduly avoid technology that can enhance the depth and vigour of their decision-making. Moreover, and as discussed above (see Boards and Directors above), it may come to be that the duty of care not only evolves to require that boards harness AI in their decision-making but also in their oversight of the company’s human-AI architecture and culture.
A question that naturally follows for the Articles is whether our understanding of the board’s duty of care should transform. AI is so complex and advancing so quickly that it’s straining traditional corporate governance accountability structures. How can boards, often made up largely of part-time, independent directors, be reasonably asked to oversee AI systems that are “iterating on products, services, and processes in real time?” Does the law as it stands reflect “unsustainable assumptions” about directors’ capacity to scrutinize AI outputs? The Article makes the point that, “much like lawyers, engineers, and other specialists, AI tools operate in ways that are difficult for outsiders to understand, much less second-guess.” They suggest that the board’s due diligence defence “ought to evolve to allow for reasonable, good faith reliance on AI outputs.” This would involve shifting the duty of care and business judgment rule analysis to focus less on the particular AI-assisted decision at issue and to focus more on the attention paid to the “design of the systems through which decisions are taken.” In other words, the Articles suggest that boards “should be able to rely on AI for smarter decisions” but should also “retain responsibility for the intentional and informed design of the systems through which such decisions occur."
Concluding Comments – Fiduciaries and Culture are Inherently Human in Nature
AI is highly polarizing. Reactions to it are often strong and can tend toward either a mix of distrust and anxiety or a mix of fascination and excitement. Impulses toward either of these extremes should be tempered. The focus should always be on objective, dispassionate and judicious analysis of AI and its implications. This applies as much to discussions of corporate governance as to anything else.
The Articles arguably lean toward the fascination and excitement side of the spectrum in several respects. One is the suggestion that the duty of care and business judgment rule should shift toward focusing less on individual board decision-making and more on the AI systems boards put in place to conduct decision-making. However, the duty of care as it stands is a flexible concept. Moreover, and practically speaking, shifting the board’s duties towards the “design of systems” would arguably drift into territory that is more appropriately the jurisdiction of management. Perhaps most importantly, we view the concept of a fiduciary as inherently human in nature.
There’s no doubt that AI is changing the expectations put on directors by their duty of care. This will in turn require ever-greater AI literacy so that directors are sufficiently informed to ask the hard questions of management that their duties require. But boards should continue to be able to rely on the advice of experts – in this case AI experts – as they always have. It must also be appreciated that directors have always been expected to bring their particular skills and experience to the table, rather than becoming experts in every area of the business. The more balanced way forward may therefore be AI-focused committees rather than board-wide AI expertise, much as boards have traditionally relied on audit, risk, and nominating and governance committees to provide concentrated oversight in complex areas. This may be a more reasonable approach than expecting all directors to develop deep expertise in AI, while continuing to maintain proficiency in their existing areas of knowledge and experience. Regardless, the focus must always remain on the application of human judgment and the hard rule that the duty of care can’t be delegated, whether to AI or elsewhere. For directors this will also mean that greater use of AI cannot be allowed to erode their critical thinking skills and other innately human talents.
This brings us to an aspect of the Articles that we think carries particular force: the notion that corporate culture will be an ever-greater competitive advantage in the AI age and that executive leadership will increasingly shift from “approving” to “coaching” and “challenging”. We generally agree that divides will increasingly emerge between companies whose culture facilitates quicker adjustment and those whose corporate cultures are more resistant to change. But this only highlights that culture and the faculties necessary to foster it – from emotional intelligence to motivational skills – are once again distinctly human in character. AI is ultimately just another tool available to boards and management. It is people who set priorities, cultivate trust, architect the values and drive the behaviors that make up the unique human fabric of an organization.
[1] See T. Chamorro-Premuzic, “How C-Suite and Board Roles Are Being Reshaped Around AI” Harvard Business Review (June 8, 2026); M. Petrin, “When AI Hurts: Corporate Responsibility for Agentic AI” (2026) 40 Harvard Journal of Law & Technology (forthcoming); D. Sarro and E. Waitzer, “Coming Disruptions in Corporate Law” (2026) 71 Canadian Business Law Journal 10; M. Petrin, “AI Accountability from the Outside In: Corporate Liability, AI Governance, and Managerial Duties” (May 2026) European Corporate Governance Institute, Law Working Paper N 935/2026.
[2] For certainty, we do not present the analysis in the Articles as our own. All opinion and insight in the Articles are those of the Articles’ authors (see footnote 1). The exception is the Concluding Comments, which is our analysis and closing thoughts, including where we push back on some of the Articles’ commentary.
[3] We do not suggest that the Articles represent any sort of consensus among business and legal experts. There is, of course, a wide variety of expert opinion as to what changes AI will bring going forward.
