Your Board Needs an AI Strategist · Not Another Director

AI risk is showing up in more board disclosures every year, but director AI expertise lags far behind. Instead of recruiting another board member, we argue for adding a strategist in an advisory chair, one without a vote, a vendor relationship, or a board seat to fill.
Board members seen from behind at a long boardroom table at golden hour, with one empty chair outlined in glowing blue beneath a rising blue bar chart.

Every board has someone who can read a balance sheet. Who reads your AI plan?

On November 10, 2026, Stanford’s business school published findings from interviews with 20 directors of public and private companies in highly regulated industries. Asked how ready their boards were to oversee AI, the directors gave a mean rating of 4.7 on a 10-point scale, and 70 percent said their boards had not clearly defined who is responsible for governing it. One of the researchers, Karen Ebert Matthys, said “most boards don’t have a great picture” of the AI their companies are rolling out. Twenty interviews is a small sample, and regulated industries are not every industry. Still, the finding matches what larger surveys keep showing and what we see in mid-market boardrooms.

The usual advice is to recruit a director who knows AI. Our view is different. Your board does need someone who can read the AI plan the way your audit chair reads the financials. It probably does not need another director to get that. It needs a strategist in a chair, and the chair does not have to be a board seat.

Your board does need someone who can read the AI plan the way your audit chair reads the financials.

 

Finance Got a Seat by Rule. AI Has Not.

Boards did not always have a financial expert either. After the accounting scandals of the early 2000s, Congress passed the Sarbanes-Oxley Act, and in January 2003 the SEC adopted a rule requiring public companies to disclose whether the audit committee includes at least one “audit committee financial expert” and, if it does not, to explain why. No rule made private companies follow. In our experience most did anyway, and today it is hard to find a mid-market board or ownership group without a CPA, a former CFO or a banker at the table.

Nothing like that exists for AI. The Conference Board reported that the share of S&P 500 companies disclosing AI as a risk in their annual filings climbed from 12 percent in 2023 to 72 percent by late 2025. Board-level AI expertise has not kept pace with that climb. Disclosed expertise undercounts what directors actually know, but the gap between a risk flagged on paper and a director qualified to read it is still hard to miss.

Private boards feel it too. When the National Association of Corporate Directors polled 76 private-company directors in June 2026, 74 percent put AI among the top issues on their board’s agenda for the next quarter, more than any other topic. It is a small poll, but AI led by a wide margin.

 

Why the Seat Stays Empty

If the need is this visible, why not just add the director? Because when directors are asked what they want from their next appointment, AI is not at the top of the list. In the 2026 edition of What Directors Think, a survey of more than 200 public-company directors fielded in the fall of 2025 by Corporate Board Member and the research arm of Diligent, a board-software vendor, AI expertise ranked fourth among the attributes directors wanted most in their next colleague, named by 28 percent. Industry expertise led at 42 percent, followed by financial or general business expertise at 34 percent and CEO or C-suite experience at 31 percent. In the same survey, only 8 percent said their board had strong AI expertise. Directors know the gap is there. They still want an operator in the seat.

The Conference Board found the same reluctance from another angle: in its surveys of governance executives at mostly large US companies, only a minority say their boards are adding directors specifically for AI or technology expertise. The respondent counts behind that question run small too, so treat any single figure as a signal, not a census.

The lawyers explain why. In an April 15, 2026 memo, the law firm Debevoise & Plimpton gave three reasons a designated AI director can disappoint. The pool of people with both deep AI knowledge and the qualifications to serve as a public-company director is limited. Few companies depend on AI so heavily that the board needs an AI expert on it. And the rest of the board may stop pushing: in the memo’s words, “other directors may defer excessively to an AI expert”. We agree with all three, and for mid-market companies we would add a fourth. In our experience seats rarely open, and when one does, the owner, the sponsor or a nominating committee usually decides who fills it. A search can run for months while the AI plan is approved without a reader.

 

Add a Chair, Not a Director

Here is what boards are doing instead, and we think the instinct is right. In that same Conference Board research, more governance executives said their boards were engaging outside advisors or consultants to build their understanding of AI than said their boards were adding directors for it. The Debevoise memo explains the footing. Under the corporate law of most states, it notes, directors may rely on experts who were chosen with reasonable care, and technical advisors can help a board assess the risks of specific AI uses.

So give that advisor a chair. An advisory seat is a different thing from a directorship. As the law firm Bradley explains in an October 2025 alert on advisory boards for private companies, advisory members are non-fiduciary and have no voting power. That is the point. You do not amend the bylaws, run a search or change who votes. Your directors keep every decision and every duty. What changes is that someone at the table has read the AI plan, knows what it should contain and will say what is missing. The arrangement can start this quarter and run month to month.

In our view the chair has four jobs across a year. Read the AI plan before the board does and say what is missing. Sit in when the board reviews it. Check each quarter whether the projects moved the numbers they promised. And tell the board plainly when a vendor’s pitch or a competitor’s move should change the plan.

Picture a $150 million building-products distributor with a five-person board: the founder, her CFO, an outside banker, a retired industry executive and a family shareholder. Management brings a 2027 AI plan with nine projects and a seven-figure budget. The banker tests the payback math. Nobody asks which projects depend on data the ERP cannot supply, which ones a competitor can copy by buying the same software, or who answers if an AI agent quotes a customer the wrong price. Now add one chair. The advisor reads the plan a week ahead, sends the board three questions and sits in for the AI item. The plan that comes out has five projects, an owner for each and a number the CFO agrees to report every quarter. The figures are illustrative. The pattern is one we see often.

 

A Strategist, Not a Specialist

Notice what the chair is for. Most board talk about AI is about risk, and that work matters; we have written about what your board will want to see on AI governance and how to prove it. But a chair that only polices risk wastes half its value. The larger question for a board is growth.

MIT’s Center for Information Systems Research studied 2,788 US public companies with more than $1 billion in revenue and reported in 2025 that the 26 percent whose boards met its test for digital and AI savvy averaged a return on equity 10.9 percentage points above their industry, while the rest averaged 3.8 points below. Those are large companies, the savvy-board companies were also much bigger than their peers, and the link is an association, not proof of cause. It still points the same way as our experience: boards that can treat AI as a growth question get more from it than boards that can only ask whether it is safe.

That is why we say strategist. The person in the chair should be an operator who has run technology for a business and answered for the results, someone who can stand on both sides of the gap between your technology and your boardroom. They should treat AI and IT as one plan, because AI built on a platform that cannot carry it stalls. They should push past efficiency, since running the same tools as your competitors makes you cheaper, not different, and they should ask where the AI plan shows up in earnings. And they should have nothing to sell you. Bradley’s alert warns against putting current vendors in advisory seats, and we would hold an AI advisor to the same standard of independence.

They should have nothing to sell you.

 

What This Will Not Do

An advisory chair does not carry the board’s duty. The advisor has no vote and no fiduciary role, so your directors still own every decision. How far they may rely on an outside expert depends on your state, your entity type and how the advisor was chosen, so have your counsel paper the role, including confidentiality. This is not legal advice.

It will not fix a missing plan. If nobody in management owns AI, the first thing a good advisor will tell the board is exactly that, and the fix is a leader, not a chair. It does not replace director education either; every director still needs enough fluency to ask a second question. And if AI is your product and not just a tool you use, you may be one of the few companies that should recruit the director.

The evidence has limits, and we have tried to show them. Most of the surveys above cover large public companies, several rest on small samples, and none measured advisory chairs at mid-market firms. That last step is our judgment, drawn from the boards and owners we have advised.

 

Where to Start

Put one question on the agenda of your next board meeting, before the 2027 plan is approved: who at this table is qualified to challenge our AI plan, and what did they challenge last time? Write down the answer. If it is a name and an example, you are ahead of most boards. If it is silence, you have found your empty chair, and you do not have to wait for a seat to open to fill it.

If you want an outside read first, an assessment will show where your technology and AI plans stand, and CIO IQ gives your board an experienced CIO’s advice at $1k per week. You can also reach us through the contact form on our about page.

No open seat? Add a chair.

No open seat? Add a chair.

More from our blog

Your Board Needs an AI Strategist · Not Another Director

Your Board Needs an AI Strategist · Not Another Director

AI risk is showing up in more board disclosures every year, but director AI expertise lags far behind. Instead of…
Your AI Dividend Can Fund a Reputation · A Role-Model Layoff

Your AI Dividend Can Fund a Reputation · A Role-Model Layoff

An AI-driven layoff can be announced like any other, or funded differently, with the savings paying for a real bridge…
U.S. Cyber Insurance Stopped Getting Cheaper · What's Negotiable?

U.S. Cyber Insurance Stopped Getting Cheaper · What's Negotiable?

U.S. cyber insurance rates finally stopped falling, but claims keep rising. We walk through what insurers now treat as standard,…