The argument in one sentence

The move every board now faces is from vibe governance to provable oversight, and AI is at once the stress test the episodic board fails and the technology that finally makes the engaged board affordable.

Two facts, arriving together

AI is accelerating strategic change beyond the cadence of episodic oversight. At the same time, AI-assisted research is sharply reducing the cost of keeping directors thickly informed. The result is the augmented board: not a board governed by AI, but directors who use it to become more deeply informed while keeping individual judgment and collective accountability.

The engaged board is not a new idea. It was mapped twenty years ago, and almost nobody arrived, because being thickly informed took more time than a part-time director has and the information ran through management. That was never a character flaw. It was a cost structure, and the cost structure just broke.

The augmented board, defined

Call it the augmented board: not a board run by AI, but a board with four properties. Common evidence, independent interpretation, timely escalation, collective judgment.

  • A common factual base. One verified set of facts, so nine directors do not arrive with nine versions of them.
  • AI-assisted but independently exercised judgment. The model is a research assistant, not an oracle. The director still checks the facts and owns the decision.
  • Information that reaches the board when material facts change. A route for a material change to reach directors in days, not at the next quarterly meeting.
  • Collective decisions supported by evidence. The interpretive layer stays individual, so the audit chair, the strategist, and the operator each read the same facts through a different lens.

Why episodic oversight fails now

The problem is not that the world changes between meetings, which was always true. It is that the rate has crossed a threshold. In one week in February 2026, the SaaSpocalypse selloff wiped out more than a trillion dollars of software value, not on missed earnings but on the fear of what AI agents would do to incumbents. The market can now reprice a whole business model on what it believes AI will do, before the damage appears in reported results.

A large-company board sits as a body for roughly 48 hours a year, about two percent of a working year. That sampling rate was fine when the business world turned over in years. The board calendar is annual. The company is now weekly. A decision that arrives after execution is not governance. It is audit.

The 83-to-3 gap

More than eight in ten of America’s largest companies now say the AI risk is real. Fewer than three in a hundred of their directors disclose any expertise in it (The Conference Board, April 2026). The obvious fix is the wrong one. Recruiting a single AI director creates a key-person risk that ISS calls the lone-expert vulnerability. Competence in a fast-moving technology is a wasting asset. It has to be distributed across the board and refreshed on a schedule.

The legal spine: Caremark, without a new duty

AI creates no separate fiduciary duty of AI oversight, and no court has recognized one. It changes how a board does the work it already has. Delaware’s Caremark line requires a good-faith, board-level system for monitoring compliance and attention to the red flags it surfaces, with particular force where compliance is mission-critical. Where consequential AI is embedded in that activity, AI-related reporting can become part of that system, and the uncomfortable question is why there was no route for red flags to reach the board between quarterly meetings.

A documented, good-faith information system strengthens the defense and makes a claim of conscious inaction harder to plead. That is protection a board earns, not a promise it will never be sued. As Kai Liekefett of Sidley Austin puts it: when directors cannot show they saw a major shift coming, the activist campaign writes itself, and on AI the engaged board is the best defense there is.

The Disclosure Mirror

The investor’s task reduces to a mirror. Flip the questions an investor asks in engagement season, and they are the questions a company must be able to answer, should disclose when material, and a plaintiff’s lawyer may later seek in discovery. The questions investors ask and the evidence a company must be able to show are the same list, seen from opposite sides.

What corporate secretaries worry about

Three objections arrive quickly, and in each the corporate secretary is the enabler, not the obstacle. Discoverability is answered by records discipline and a clear line between briefing and deliberation, designed in rather than retrofitted after the first records demand. Confidentiality is answered by a board-approved tool: only 2 percent of companies give directors a sanctioned one, even as most directors already use AI in the role, and that gap between sanctioned and actual use is the real exposure. The worry that an augmented board freelances into management’s lane is answered by design: the architecture changes, not the line of authority. It reads more. It does not manage more.

The tell: vibe governance

There is a fast way to tell whether a board is driving or just holding the wheel: ask who owns AI risk and what evidence shows the controls work. If the answer is a list of comforting inputs, a policy, a framework, a training, a tool, that is what Nora Denzel of AMD calls vibe governance. Reassuring and empty, because no one owns the outcome. Her test holds: if it will only probably hold up when a regulator or a plaintiff’s attorney comes knocking, you do not have governance, you have hope.

Citation

Maciejko, Robert. The Augmented Board: AI, Caremark, and the End of Episodic Oversight. SSRN, August 4, 2026. Available at ssrn.com/abstract=7157338. This is the director’s cut of the author’s article in Insights: The Corporate & Securities Law Advisor, expanded for boards and directors.

Go deeper