Insurance Insider – AI liability exclusions and demand could spike after 1.1 2027: ibott’s Moore

“I think being silent on this topic will not last very long,” he told Insurance Insider.

The vast majority of insurance policies today are silent on AI liability, but 1 January 2027 could be a “big tipping point” of AI exclusions and demand both increasing, Chris Moore, Apollo ibott commercial president, told Insurance Insider.

Some markets are not really making changes to their products, whether it’s putting an exclusion or an affirmative type of coverage in place, the executive said in an interview with this publication.

In fact, for most commercial insurance policies, “it’s highly unlikely that they even have a definition of what artificial intelligence is inside the form,” he added.
These instances fall into the category of silent AI, which Moore feels is a dangerous position to be in as an insurer.

“I think being silent on this topic will not last very long,” he said. From a client perspective, insurers have the “promise to pay” when they have a loss they believe is covered. “If [the policy] doesn’t respond to that, you’ve broken that trust,” Moore added.

And that could be a big miss for a market that’s bound to grow. “I certainly think AI, whether it will be its own vertical or not, will be the biggest opportunity for the insurance market in the next decade, undoubtedly,” Moore said.

Apollo ibott, also known as Syndicate 1971, started by providing insurance for autonomous vehicle testing in San Francisco. Today, a big chunk of its clients are US-based tech companies in the sharing economy and mobility space, where historic data was scarce a decade ago when the team started working with them.

With that experience, the London syndicate currently offers general and product liability coverage for AI developers, as well as providing solutions for users of AI whose traditional insurance policy may not cover AI-related exposures.

In the broader marketplace, AI exclusions are just starting to emerge, according to Moore.

In January, Verisk published a generative AI exclusion for commercial general liability policies. The company is working on introducing agentic AI exclusions as well, this publication reported last month.

“I think there’ll be a big shift at January 1, 2027,” Moore said. “You’ll see a lot more exclusions added as the language is released and [that] will force a growing demand because people will be using AI and they’ll want real clear coverage.”

“That’s why it’s been really important for us to be developing products way before 1.1 2027 – to be ready for that.”

Moore anticipates there will probably be a period where most traditional carriers exclude AI as standard practice.

But as clients demand AI coverage, carriers will likely go in two directions: either be affirmative that AI losses are covered in their cyber or general liability form or offer a wraparound or standalone product for AI liability that can fill the gap created by exclusions.

“That remains to be seen which one of those solutions wins in the long run. I think it will probably be a hybrid of a few different approaches,” he said.

When it comes to the most sophisticated clients, Moore predicts that most of them will end up buying standalone towers, seeking more bespoke coverage in terms of what’s covered and what’s not, and how the program is priced.

A recent observation from the client end is AI developers ramping up hires of experienced risk managers, a lot of whom Moore has worked with in the mobility and tech sectors.

“I’m seeing quite a high degree of frequency of this person start a new job at an AI development company and they are really impressive, very seasoned risk managers,” he said. “They’re building out what I would say is first-class risk management teams.”

That sophistication will drive serious demand in insurance, Moore added. But at the same time, it can pose a threat for traditional insurers.

“When people ask, ‘Who is ibott’s biggest competitor in the sharing economy, autonomous vehicles, new mobility, AI liability [space]?’, I always answer the same way: it’s my clients,” he told Insurance Insider.

If an insurer can’t provide a solution that’s sustainable from a pricing perspective and is designed to cover their specific exposures, clients will in-source insurance, Moore explained.
There’s already been precedents: Uber and Lyft, which in their early days struggled to find coverage at the right price, now operate big captives. Tesla launched its own insurance company in 2019.

“That is a warning shot for me. Our clients have huge balance sheets – they can build fronts, insurance companies and create their own insurance solution,” the executive said.
“They’re not going to let insurance stop them in their journey and the progress in their technology. And I think that’s the key – insurance should be an enabler, not a hurdle.”

But when it comes to potential demand, Moore believes that the bigger need for AI liability coverage will come from users of AI.

Indeed, AI developers will buy very large liability towers. But at the end of the day, they will be a small cohort and have the balance sheet to take on large retentions.

“Without a doubt, the biggest market is the deployers [of AI] because that is pretty much every company on the planet in every industry,” Moore said, specifying healthcare, legal consulting and media and advertising as areas where he sees big demand.

In the past 12 months, the rollout of AI adoption strategies has accelerated, with a much greater proportion of businesses (63%) now having either fully operationalised or implemented AI within parts of their business.

According to Gallagher’s 2026 AI Adoption and Risk Survey, 63% of corporate respondents said they have now fully either fully operationalised or implemented AI within parts of their business, compared to 45% last year.

In Moore’s view, many companies that use AI as a tool believe their exposure hasn’t changed, and therefore, insurance coverage will not be affected, because there’s a “human in the loop” that does the final decision-making. So even if there is a negligence case, that would be “human negligence” instead of AI.

But as AI technology continues to advance, top corporate leaders could start to think that having a human in the loop may not be the best thing or is holding the company back.
“That’s where I see a huge spike in demand,” Moore said.