AI has passed a tipping point. It’s no longer just a background tool, with companies increasingly using the technology to shape strategy – with 98% of mid-market businesses exploring how AI can support high-level decision-making. And it goes further, with some companies and even government bodies augmenting human leadership by giving AI a level of authority that starts to resemble an executive role.
That might sound theoretical, but it’s happening today – and it points to a rising challenge when it comes to directors and officers (D&O) risk: when decisions are no longer made purely by people, who is actually accountable when something goes wrong?
How D&O risk is structured today
D&O risk has always been built on the simple premise that decisions are made by people, and those people can be held accountable.
As such, D&O policies are designed to protect individual directors and officers when claims arise from their actions, decisions or oversight. Be it a strategic misstep, a failure in governance, or a regulatory issue, responsibility can be traced back to a defined individual or group.
But AI starts to challenge that assumption. When decisions are influenced – or effectively made – by a system, the line between human judgement and machine output becomes harder to separate. If something goes wrong, it’s no longer obvious if the responsibility sits with the individual, the organization, or the technology itself. And that uncertainty is exactly where challenges for D&O cover begin to emerge.
Where D&O wordings start to come under pressure
-
Accountability is harder to pin down
D&O cover is built around identifiable individuals. When a decision is driven by AI, it’s less clear whether responsibility sits with the board, the individual relying on the output, or the business itself.
-
The rise of ‘digital twins’
As AI tools become more advanced, they can effectively act as digital versions of directors – analyzing information, recommending actions and influencing outcomes. That raises questions around whether these systems are simply support tools, or something closer to a decision-maker in their own right.
-
Greater exposure from AI-driven decisions
AI isn’t just changing how decisions are made, but can increase the impact when something goes wrong. Decisions made at speed, based on flawed or biased inputs, can lead to larger financial or regulatory consequences than a single human error.
-
Oversight and challenge risk
If AI outputs are treated as authoritative, there’s a risk that human oversight becomes more passive. That creates more risk if boards are seen as failing to properly challenge or validate decisions they are ultimately responsible for.
The verdict: is D&O insurance fit for today’s world?
D&O insurance isn’t suddenly out of date, but it is being asked to respond to a very different type of risk. The core principle still holds: protecting individuals for decisions made in their role. The challenge is that those decisions are no longer entirely human, and that creates grey areas around responsibility, oversight and how policies are expected to respond.
For now, much of this sits in that grey space between existing cover and emerging exposure. As AI adoption accelerates further, those questions will become harder to ignore. It’s an area we’re watching closely as it evolves, and one that’s likely to shape how D&O risk is understood in the years ahead.
Stay up to date on our latest AI content here. And get in touch if you have any questions about AI, D&O risk and CFC’s management liability cover.
Meet the author
Kate Lyes is Head of Specialty Lines at CFC. She leads the strategy and development of a broad portfolio of specialty products, working closely with brokers to deliver innovative solutions and helping businesses navigate complex and evolving risks with confidence.
