OnBoard says board AI policy key, but lags behind in adoption

OnBoard, the board governance software provider, has warned that boardroom adoption of artificial intelligence is significantly outpacing formal policy, with the majority of boards still operating without dedicated AI guidelines.

The company's sixth annual Board Effectiveness Survey, which gathered responses from more than 500 board directors and executives, found that 92 per cent of board directors now use AI for board-related work, up from around 69 per cent a year ago. Despite this rise in adoption, 60 per cent of boards have no formal AI policy in place, while 41 per cent have no guidelines at all.

"The individual is far out and ahead of IT and the organisation itself," said Rob Kunzler, Chief Marketing Officer, OnBoard. "So much of this AI adoption, whether it be by board directors or just your average employee, is embracing the tools and tech that's available to them and experimenting with it, often before some centralised function has adjudicated the way in which that technology should be used, approved or not approved."

OnBoard's survey focused specifically on board directors and executives with an interface into board work, rather than employees more broadly. Kunzler said he expected similar patterns to hold across organisations, noting that board directors may, in some cases, have greater access to commercial AI tools than junior staff, given subscription costs and procurement processes.

The survey also found that organisations with a formal AI policy at the board level reported higher board effectiveness than those without one.

"Organisations that do have a formal policy in place - a very specific policy in place at the board level, self-report that their own effectiveness increases rather than decreases," said Kunzler. "Having a policy in place is very strongly suggestive of a board operating at a higher, more strategic, and more effective level."

A striking fact, according to Kunzler, is that board directors surveyed consistently identify at least one colleague as ineffective, a finding that he said has remained stable across six years of research. Directors most commonly attribute this to a lack of preparation ahead of meetings, he said, adding that AI tools could help address the issue by giving directors, including new appointees, a fuller context on prior board activity.

Board turnover was also cited as a factor. Kunzler said newly appointed directors are often perceived as less effective for a time simply because they lack the historical context that longer-serving members have accumulated, and that AI tools built for governance could help close that gap more quickly by giving new directors fuller visibility into past board activity.

"The first step is committing at the board level that a policy itself is both useful and required, and is in and of itself best practice," said Kunzler. He said boards should then draw on existing best-practice frameworks, consult directors on the policy's content, and align it with the organisation's charter and objectives. Policies should also be revisited periodically as technology changes, he added, citing the shift among some organisations from ChatGPT toward other AI platforms like Claude as an example of why flexibility matters.

OnBoard will hold a keynote on 15 July, setting out its view of the future of board governance technology. Kunzler said the event would address an industry with an estimated five million board-led organisations across the U.S. and Canada.

"Adoption is outrunning governance, and that's why this year we believe that this gap is something that organisations should look to alleviate," said Kunzler. "Most importantly, we believe that the intentional and thoughtful use of AI can be a separator rather than a limiter of board effectiveness."