Insuring AI-related risks

13 Aug Insuring AI-related risks

Posted at 13:27h in 2025, Risk Management, Uncategorized by

Anita Byer

Businesses are adopting artificial intelligence at a staggering pace and using AI in more business functions than ever before. According to a McKinsey and Company survey, after years of minimal change, the organizational use of AI has accelerated significantly, from 33 percent in 2023 to 71 percent by mid-2024. For the first time, most respondents reported using AI in multiple business functions. But as the use of AI increases, so do the number of AI-related incidents. This means that businesses seeking to expand their capabilities with AI must also understand and protect against AI-related risks.

According to the 2025 Stanford AI Index Report, there was a record number of AI-related incidents last year and a 56 percent increase over 2023. Though the full extent of AI-related risks remains to be seen, the consequences of AI-related errors can be severe. According to Munich Re, the types of losses and claims that can result from malfunctioning AI can include:

  • Property damage, bodily injury, death (driverless cars, automated machines, industrial robots)
  • Administrative fines or penalties
  • Privacy violations (unlawful collection, retention, or disclosure of confidential, personal or identifying information)
  • Data leaks (dissemination of trade secrets or confidential information resulting in financial losses, public embarrassment, loss of shareholder or customer confidence)
  • Intellectual property infringement
  • Pure financial losses
  • Underperformance, inaccuracies, hallucinations
  • Defamation
  • Discrimination

 

The cost of AI-related errors is only expected to go up as businesses incorporate AI into increasingly vital operations. The insurance market for AI-related risks is in its infancy but developing rapidly. In fact, the Deloitte Center for Financial Services projects that global premiums for AI insurance could reach $4.7 billion by 2032.

Until the market matures, however, some businesses may be able to rely on traditional insurance policies to respond to various AI-related claims under some circumstances, provided there are no AI-specific exclusions. A cyber liability policy, for example, could respond to an otherwise-covered data breach caused by AI. An employment practices liability policy could respond to claims of discrimination resulting from AI’s biased choices. A workers’ compensation policy could respond if an employee is injured by AI-controlled machinery. This kind of coverage would not be available under policies that specifically exclude losses caused by AI, so selecting the right coverage form is crucial.

Please contact us to learn more about mitigating AI-related risk exposures.

 

Print page