The Augmented Executive

AI Oversight From the Board Seat

Artificial intelligence will change how every officer decides. The board must ensure that augmented leaders still think for themselves.

In Brief

The Problem. AI is changing how officers make decisions. The gains are real, and so are the risks: dependence on a few model providers, the erosion of independent judgment, and recommendations no one can explain.

Why It Persists. Boards treat officers' use of AI as a management matter. Delaware law now says otherwise: officers carry a duty of oversight within their areas, and the board supervises the officers.

What Boards Should Do. Make AI part of each officer's evaluation, know which decisions depend on which providers, and require an accountable person in the loop for critical decisions.

For decades, senior management has been a remarkably human affair: a blend of experience, intuition, and analysis of historical data. The chief executive scanned last quarter's results, the chief financial officer studied last year's accounts, and the chief operating officer analyzed yesterday's output. The perspective was always in the rear-view mirror. Artificial intelligence (AI) turns it forward.

AI will alter the job of every executive in the C-suite. It will not make leaders redundant; it will augment them, combining human judgment with machine-generated foresight. My argument is that this is a board matter, not only a management matter. The board selects, evaluates, and supervises the officers, and the law now expects officers to oversee the systems they rely on. The board's task is to ensure an augmented leadership team, not a homogenized one.

How Each Role Shifts

• The chief executive moves from setting strategy at an annual retreat toward continuous testing of scenarios, competitive threats, and market shifts. Strategy becomes a living process rather than a static document.
• The chief financial officer moves from financial historian toward forecaster of risk: running continuous forecasts, allocating capital as conditions change, and detecting financial anomalies before they become crises. Gartner found in 2024 that 58 percent of finance functions were using AI, and anomaly and error detection was among the most common uses (Gartner, 2024).
• The chief operating officer moves from managing supply chains toward anticipating disruption across a network of suppliers and rerouting before a shortage reaches patients or production.
• The chief information officer moves from maintaining systems toward searching for security threats before they cause harm, rather than reacting to breaches.

In health systems, the same shift reaches the chief medical and chief nursing officers, whose clinical decision support tools now shape care at the bedside. The American Medical Association (AMA) reported in March 2026 that 81 percent of physicians use AI in their practices, up from 38 percent in 2023 (American Medical Association, 2026). In life sciences, it reaches the heads of research, regulatory affairs, and pharmacovigilance.

The deeper change is in how organizations create value. Advantage will come less from a superior product alone and more from a superior way of deciding, provided the data is clean and the models are sound. Organizations that combine human leadership with AI-generated insight well should see shifts in demand sooner and move research faster. That is my expectation, not a proven result, and it depends on the risks below being governed.

Why It Is a Board Matter

The change carries a legal dimension. In January 2023, the Delaware Court of Chancery held in In re McDonald's Corp. Stockholder Derivative Litigation, 289 A.3d 343 (Del. Ch. 2023), that officers owe a duty of oversight within their areas of responsibility (Raymond et al., 2023).

The principle has since reached life sciences. In the Teligent case, decided on September 2, 2025, oversight claims proceeded against the chief executive and the chief scientific officer of a drug maker. Compliance with Food and Drug Administration requirements fell within their areas, and they allegedly failed to report known issues to the board. Claims against the chief financial officer were dismissed, because that compliance was not among that officer's core responsibilities (Bonadurer, 2026).

An officer who relies on AI in a mission-critical area must oversee it and report on it. The board, in turn, supervises the officers. Its evaluation of each executive should therefore include how that executive uses AI, monitors it, and brings problems forward.

Three Risks to Watch

Dependence on a few providers. Many AI tools rest on models built by a small number of technology companies. When an executive team's analysis runs through the same external models, the team risks becoming a local extension of an outside system whose design choices reflect the provider's goals more than the organization's. The board should know which decisions rely on which providers, and what the contracts allow.

Deskilling. Over-reliance on the machine can erode the judgment and the willingness to dissent that make an executive valuable. When a system always serves up a polished, consensus answer, contrarian views fade. The risk is now measured. In a 2025 multicenter study in Poland, endoscopists who had routinely used AI to detect polyps found fewer adenomas when they worked without it. Their detection rate fell from 28.4 percent before AI was introduced to 22.4 percent after (Budzyń et al., 2025). A leader who follows AI recommendations without scrutiny is as dangerous as one who ignores the data entirely.

Opaque reasoning. When the reasoning behind a recommendation cannot be explained, accountability disappears. If an executive cannot explain why a decision was made, the board cannot evaluate it, and the organization cannot defend it.

Augmentation Reaches the Workforce

The same principles apply below the executive suite. The anxious question about AI at work is who will be replaced. The more useful question is who will be augmented, and how well. For decades, technology codified existing processes and made them faster and cheaper. AI can also surface new approaches, freeing people for complex problem-solving and judgment.

Drug discovery shows the partnership at its most advanced. In June 2025, Insilico Medicine and clinical investigators reported a randomized phase 2a trial of rentosertib in idiopathic pulmonary fibrosis. Both the drug's target and the molecule itself were identified using generative AI. The drug was safe and well tolerated over 12 weeks, and the highest dose showed an early signal of improved lung function that requires larger trials to confirm (Xu et al., 2025). The scientists did not disappear from that work. They chose the disease, judged the candidates, designed the trial, and interpreted the results.

The partnership also works in daily operations. Consider an illustration: a forecasting model flags an unexpected drop in clinic volumes; a manager supplies the context, such as a local holiday or a competitor's new site; the model is updated with that context, and its next forecast improves. The advantage comes from combining people and machines on problems neither would solve as well alone.

That partnership depends on three commitments the board should expect to see.

First, critical decisions keep a human in the loop: the final call rests with an accountable person, not with a system. California has now made that a legal requirement for clinical care. Its Assembly Bill 1979, effective January 1, 2027, requires facilities to ensure that a licensed provider uses independent judgment when AI informs care (Montano & Tran, 2026). The cost of losing that discipline is visible in litigation. In Estate of Lokken v. UnitedHealth Group, plaintiffs allege that the insurer used an algorithm, nH Predict, in place of physicians to cut off post-acute care and required employees to follow it; UnitedHealth denies using the tool. In February 2025, a federal court allowed the breach of contract and good faith claims to proceed (Estate of Lokken v. UnitedHealth Group, Inc., 2025). The allegations are unproven, but they show what a board will be asked if "human in the loop" exists only on paper.

Second, managers learn a new role. A manager becomes less a supervisor of tasks and more a coach for both the team and its tools, teaching each to work better with the other.

Third, the organization invests in continuous training, so that staff move from performing tasks to managing tools and questioning outputs. An advantage built on how people and tools work together is embedded in the organization, not only in its technology, and it is difficult for competitors to copy.

The Objection: This Is Management's Business

The strongest objection is that how officers use their tools is a management question. Boards do not evaluate which spreadsheet a chief financial officer prefers, and asking about AI in every evaluation breaks the rule of "nose in, fingers out."

I agree that the board should not choose tools or review prompts. The question is whether an officer can still exercise and explain independent judgment, and whether that officer oversees the AI in a mission-critical area. Since McDonald's and Teligent, those are questions of officer duty, and evaluating officers is the board's work. A spreadsheet does not make recommendations. An AI system does.

Leading by Cross-Examination

The art of management in this century will be knowing when to trust the machine and when to trust the person. It calls for leaders who need not be coders but who are literate in data, probability, and the logic and limits of a model. Leadership becomes interrogation, not only interpretation. A good leader cross-examines an AI recommendation, probing for hidden assumptions and bias, as counsel would question a witness.

When I evaluate an executive's judgment, I look for the moment the person disagreed with the data and could explain why. Joseph Schumpeter called the process by which new methods displace old ones "creative destruction" (Alm & Cox, n.d.). That process is now reaching the corner office itself. The challenge for boards is to ensure that what emerges is an augmented human team, not a uniform one in which dissent is treated as an error to be corrected.

For the Board

• Which executives rely on AI in mission-critical decisions, and how does each oversee it?
• Does our evaluation of each officer include how that officer uses, monitors, and reports on AI?
• Which of our key decisions depend on the same external model providers, and what do the contracts allow?
• Where do our critical decisions keep an accountable person in the loop, and what evidence shows that the person can and does override the system?
• How are we measuring and preserving the skills our clinicians and executives need when the tool is unavailable?
• When did an executive last disagree with an AI recommendation, and what did we learn?

To brief your directors on officer oversight of AI before the annual executive evaluation, book a 30-Minute Board Briefing with Avancer.

References

Alm, R., & Cox, W. M. (n.d.). Creative destruction. In The concise encyclopedia of economics. Library of Economics and Liberty. Retrieved October 21, 2026, from https://www.econlib.org/library/Enc/CreativeDestruction.html

American Medical Association. (2026, March 12). AMA: AI usage among doctors doubles as confidence in technology grows [Press release]. https://ama-assn.org/press-center/ama-press-releases/ama-ai-usage-among-doctors-doubles-confidence-technology-grows

Bonadurer, J. (2026, February). Delaware court denies dismissal of Caremark oversight claims in Teligent case, highlighting board and officer compliance obligations. Jenner and Block. https://www.jenner.com/en/news-insights/publications/delaware-court-denies-dismissal-of-caremark-oversight-claims-in-teligent-case-highlighting-board-and-officer-compliance-obligations

Budzyń, K., Romańczyk, M., Kitala, D., Kołodziej, P., Bugajski, M., Adami, H. O., Blom, J., Buszkiewicz, M., Halvorsen, N., Hassan, C., Romańczyk, T., Holme, Ø., Jarus, K., Fielding, S., Kunar, M., Pellise, M., Pilonis, N., Kamiński, M. F., Kalager, M., . . . Mori, Y. (2025). Endoscopist deskilling risk after exposure to artificial intelligence in colonoscopy: A multicentre, observational study. The Lancet Gastroenterology & Hepatology, 10(10), 896–903. https://doi.org/10.1016/S2468-1253(25)00133-5

Estate of Lokken v. UnitedHealth Group, Inc., No. 23-3514 (JRT/DJF) (D. Minn. Feb. 13, 2025) (memorandum opinion and order). https://litigationtracker.law.georgetown.edu/wp-content/uploads/2023/11/Lokken_2025.02.13_MEMORANDUM-OPINION-AND-ORDER.pdf

Gartner. (2024, September 11). Gartner survey shows 58% of finance functions using AI in 2024 [Press release]. https://www.gartner.com/en/newsroom/press-releases/2024-09-11-gartner-survey-shows-58-percent-of-finance-functions-use-ai-in-2024

Montano, M., & Tran, S. (2026, October 8). Newsom approves 3 health care AI bills, vetoes 2. California Health Care Foundation. https://chcf.org/blog/2026/10/08/newsom-approves-3-health-care-ai-bills-vetoes-2

Raymond, F. D., Lidondici, C. M., & Schiltz, T. C. (2023, February 23). Delaware Court of Chancery expands duty of oversight to corporate officers. Faegre Drinker. https://www.faegredrinker.com/en/insights/publications/2023/2/delaware-court-of-chancery-expands-duty-of-oversight-to-corporate-officers

Xu, Z., Ren, F., Wang, P., Cao, J., Tan, C., Ma, D., Zhao, L., Dai, J., Ding, Y., Fang, H., Li, H., Liu, H., Luo, F., Meng, Y., Pan, P., Xiang, P., Xiao, Z., Rao, S., Satler, C., . . . Zhavoronkov, A. (2025). A generative AI-discovered TNIK inhibitor for idiopathic pulmonary fibrosis: A randomized phase 2a trial. Nature Medicine, 31(8), 2602–2610. https://doi.org/10.1038/s41591-025-03743-2