At the recent 2025 Corporate Board Member conference in Washington, DC, one small MIT study sparked a big discussion which Jennifer Robertson, Managing Director of Board Matters has reflected on. Here are her thoughts. 

“The study found that when people use AI (like ChatGPT) from the very start of a writing task, they tend to outsource their thinking entirely. The result? Their own cognitive engagement drops off when using AI.

In my view, this is a huge red flag for governance.

For directors, reading the board pack isn’t just about staying informed. It’s about identifying nuance, sense-checking, and bringing independent thought into the room. AI can absolutely help… but only after directors have done their own mental heavy-lifting.

I am convinced that directors who show up having read, reflected, and formed an initial view will be far more effective at probing assumptions, challenging themselves and the management team. AI might synthesise content, but it can’t replicate context, lived experiences, or that all important ‘gut check’ moment. It doesn’t (yet) know some of the tell-tale signs of a looming disaster (or where the bones are buried!)

There’s also a risk in overconfidence. A slick AI summary might feel like mastery, but it could gloss over the detail that changes the story. It is in those key details where governance risks hide.

At the end of the day, governance is about applying wisdom and not about absorbing information. Tools like AI are incredibly powerful, but they are just that: tools! They don’t replace a director’s role in thinking deeply, questioning thoroughly, and engaging meaningfully.

My take? Read first (do the hard work!) then – and only then – start interrogating with AI to see what you may have missed or issues you wish to explore further. That’s how humans will be board-ready in an AI-enabled world.”