Half of firms flag AI errors as insurers lag coverage

The report finds liability often falls on companies deploying tools, not developers.

Businesses are facing growing legal and financial risks as they adopt artificial intelligence faster than insurers can adapt, according to a Gallagher Re paper.

Written by Katrina Hill and Ed Pocock, the report says many AI-related losses fall outside traditional policies such as cyber, professional liability and product liability insurance. 

It highlights risks including false outputs, biased decisions, model failures and contaminated data, which can lead to legal claims even without a cyber attack or human error. 

Legal cases linked to generative AI are rising quickly. 

Businesses are also increasingly concerned. In a survey of 1,250 companies, 57% identified AI errors and misinformation as a key risk, whilst over half cited legal, reputational and data privacy concerns. 

The report says responsibility often falls on companies using AI rather than those that build it.

It adds that existing insurance cover is limited. Cyber policies may cover AI-related attacks, but not liabilities caused by AI outputs such as defamation or copyright disputes. 

Insurers are beginning to respond with specialised products, but the report says a gap remains. 

It warns that as AI use expands, businesses may be left exposed unless coverage evolves more quickly.