What Education Boards Are Saying About AI: Key Findings from AGB OnBoard’s Board Effectiveness Survey

The findings are in, and for higher education governance leaders, they raise questions worth taking seriously.

OnBoard’s sixth annual Board Effectiveness Survey captures governance data from hundreds of board leaders across sectors each year, including higher education. This year, AI emerged as a dominant theme, not as a distant future topic, but as a governance challenge boards know they need to address and aren’t sure they’re equipped for.

Here is what the data shows across the governance landscape, and what it means for higher education boards specifically.

AI adoption is outpacing AI governance

Across the sectors surveyed, AI adoption has moved fast: 92% of board directors reported using AI for board work in the past six months, up from 69% just a year earlier. Directors are using it to summarize board books, draft minutes and agendas, research topics, and prepare for meetings. But far fewer boards have formally addressed AI governance at the board level: documented oversight responsibilities, formal committee assignments, or structured reporting on AI risk.

That gap matters. Accreditation bodies are beginning to ask how institutions document decision-making processes that involve AI. State legislatures in several markets are introducing oversight requirements. Faculty and donors are watching how institutions govern technology that touches their community.

The board’s job isn’t to manage AI. It’s to ensure management has a framework for doing so, and that the board is kept meaningfully informed.

Trustee preparation is the common variable

The survey also revealed a consistent pattern: boards that reported strong governance outcomes were those where trustees felt genuinely prepared walking into meetings. Preparation quality correlates directly with governance quality, and preparation quality depends on the tools and infrastructure supporting it.

The policy data makes the governance stakes concrete. Boards with an enforced AI policy rate themselves 88% effective. Boards with no policy rate themselves 55% effective. That is a 33-point gap. Only 6% of boards surveyed have an enforced policy in place. The gap isn’t intent. It’s structure.

What effective boards have in common

The policy maturity picture is stark across every sector surveyed: 38% of boards have no AI policy at all, 25% are still developing one, and only 6% have reached an enforced stage. For higher education boards already under accreditation and legislative scrutiny, that distribution is an institutional risk, not a future consideration. The boards that move now will have documented frameworks when it matters. The boards that wait will be building them under pressure.

OnBoard AI: Built for the Governance Demands of 2026

OnBoard AI addresses the preparation and documentation gaps the Board Effectiveness Survey identified, giving boards the structure to meet with greater focus, govern with greater precision, and produce records that hold up when accreditors, legislators, or legal counsel come asking.

For higher education boards navigating a demanding governance environment, that infrastructure is not optional. It is the foundation.

On July 15, OnBoard is addressing this gap directly. CEO Marc Huffman will make the case for why AI in governance should extend human judgment, not replace it, the same distinction this data draws between adoption and oversight. Chief Product Officer Tim Adair will walk through what that looks like in practice: AI that works from your board’s actual record, inside the permissions your governance framework already defines. And Chief Customer Officer Anusha Srijeyanathan will show what it means for that AI to compound over time, so institutional memory doesn’t walk out the door when a trustee rotates off.

Explore OnBoard AI Assist | Register for the July 15 Keynote | Download the full Board Effectiveness Survey