Will AI Need Its Own Insurance Policy?

As artificial intelligence (AI) continues to dominate headlines, the insurance industry is grappling with a pressing question: will AI-related risks eventually require their own standalone insurance policies?

With a growing number of businesses reporting AI adoption in some form, but only a fraction expressing confidence in how those risks are being managed, some industry leaders believe history may be repeating itself. Much like cyber liability, which was once covered only through general liability or E&O policies before evolving into a specialty line, AI risk may be on a similar trajectory.

At the Women in Insurance Summit, leaders from across the industry discussed the evolving risk landscape. “I saw a stat that 79 percent of businesses are using AI in some way right now, and that 32 percent of businesses feel confident that it's being risk managed or it's being risk transferred. So there’s definitely a gap there,” said Kate Della Mora, CEO of CFC Underwriting Inc. for Canada.

The idea of a standalone AI insurance policy isn’t just speculative – it could mirror a pattern the industry has seen before. When cyber risk first emerged, it was initially viewed as a component of broader coverage lines. Over time, however, as claims data accumulated and the scope of digital exposures grew, it became clear that cyber required dedicated underwriting models, policy language, and risk mitigation strategies.

A similar inflection point may be approaching for AI. Right now, most AI-related exposures are presumed to fall under existing policies – whether through general liability, cyber, or E&O coverage. But as companies deploy AI more broadly and in increasingly critical business functions, insurers may begin to see claims patterns that aren’t adequately addressed by traditional coverage.

Della Mora noted that clients are increasingly requesting affirmative AI coverage language, seeking clearer assurances about how their AI-related risks are being addressed. Whether or not this leads to a dedicated AI policy will depend on several factors: how quickly adoption scales, the types of claims that emerge, and whether existing policy language can adapt.

If current coverages prove insufficient in real-world scenarios – as was the case with cyber – the industry may be compelled to carve out AI as a specialty product in its own right.

Understanding AI maturity is crucial. Vinita Jajware-Beatty, president and board chair of the Toronto Insurance Women's Association, said that underwriting AI risks isn’t just about identifying the presence of the technology – it’s about evaluating how well it's governed. From an enterprise risk management (ERM) perspective, insurers will need to assess an organization’s AI maturity, including the internal controls, oversight mechanisms, and ethical guardrails that are in place.

This kind of governance scrutiny will help underwriters determine not only the level of exposure but also whether standalone AI coverage is truly warranted. Jajware-Beatty also pointed out that brokers working with AI-heavy firms – particularly in sectors like proptech and fintech – should already be considering how those risks intersect with existing technology errors and omissions (tech E&O) policies.

From the broker's point of view, any movement toward standalone AI coverage won’t just be driven by demand from clients – it will also depend on how insurers respond to emerging risks. Amanda May, vice president at McCAM Insurance Brokers, emphasized that brokers will be watching closely to see which carriers are willing to step up.

As AI-related claims begin to surface, especially if they are denied under existing policy terms due to ambiguous or absent AI-related language, brokers will be eager to identify insurers that proactively address the coverage gap. May sees this as an opportunity for product innovation and leadership in the market.

May noted that this moment could be especially pivotal for insurers’ product development teams. Those that act early to offer targeted AI coverage – or at least incorporate clear AI clauses into existing products – will likely become go-to partners for brokers looking to protect clients navigating new technological frontiers.

If we see that there are claims being denied and it's attributed to AI clauses, it's going to be a question of who's going to step forward and say: ‘We're going to fill that gap for you’.