The Development
Board Intelligence this week published its Board Value Index for Summer 2026, and the most consequential finding is not about a competitor, a regulator, or a market. It is about the boardroom itself. Two thirds of directors, 66 percent, now report using artificial intelligence in the course of their board work. Fewer than one in four, just 22 percent, sit on a board that has adopted any policy governing that use.
The same Board Intelligence study found that 86 percent of directors believe overly rigid or inconsistent decision frameworks contributed to a delayed, rushed, or poor decision in the last six months. Tool adoption is running well ahead of the governance scaffolding meant to contain it.
In plain terms, the technology your board is charged with overseeing has already taken a seat at your own table, and in most cases it arrived without rules.
Why It Matters to the Board
Boards have spent two years debating how to oversee management’s deployment of AI across the enterprise. The Board Intelligence data reframes the question. The first AI governance gap a director should worry about is not in the finance function or the customer service queue. It is in the board’s own workflow.
When a director pastes a confidential draft into a consumer chatbot to summarize it, material nonpublic information may leave the company’s control. When AI generated analysis informs a board decision and no one has validated the source, the duty of care is exposed. Oversight credibility erodes quickly when the overseer operates by a looser standard than the operations it supervises.
The Risk If You Wait
The liability exposure is concrete. Under the Caremark line of Delaware decisions, directors must make a good faith effort to implement and monitor a reporting system for mission critical risks. AI now qualifies as exactly that kind of risk for most public companies, and a board that uses the technology daily while documenting no policy is building its own record of inattention.
Waiting also compounds. Every month without a director AI usage policy is another month of board materials, minutes, and deliberations potentially touched by ungoverned tools, with no audit trail to reconstruct what was relied upon and what was not. Rebuilding that history after an incident is far more expensive than governing it in advance.
What Other Boards Are Doing
The leading boards have moved. The number of S&P 500 companies with a dedicated board level AI oversight committee tripled during 2025, and roughly 40 percent of Fortune 100 boards now assign AI to a named committee, up from 11 percent in 2024. The structural response is well underway.
What separates the strongest boards is that they have turned the lens inward. They have adopted a short director AI usage policy that defines approved tools, prohibits entering confidential or material nonpublic information into unapproved systems, and requires disclosure when AI materially shapes analysis presented to the board. They treat their own conduct as part of the governance perimeter, not outside it.
The Governance Question
Every board should be able to answer one question at its next meeting. If a regulator or a plaintiff asked how this board governs its own use of artificial intelligence, what document would we produce?
If the honest answer is that no such document exists, the board has identified its most immediate AI governance gap, and it is one entirely within its power to close before the next session.
Intelligence Bottom Line
AI adoption inside the boardroom has outpaced AI governance inside the boardroom by a factor of three. The fix is neither costly nor technical. A one page director AI usage policy, ratified and minuted, converts an unmanaged fiduciary exposure into a documented control. Boards that act this quarter will govern from a position of credibility. Boards that wait will be explaining the gap later, on someone else’s timeline.