Authors: Katrina Hill Ed Pocock

Smart Systems, Blind Spots: Rethinking Insurance for the AI Era, written in association with MIT and Testudo, explores how the rapid deployment of artificial intelligence (AI) has outpaced the insurance market's ability to address the unique risks it introduces. As organizations integrate generative and non-generative AI into customer interactions, decision-making and core operations, they may inadvertently expose themselves to a class of AI-native liabilities that traditional insurance policies may not fully recognize or respond to.

These liabilities differ from conventional risks like cyber events, professional errors or product defects. Instead, they stem from AI-specific failures such as hallucinated outputs, flawed automated decisions, discriminatory behaviors, model drift and compromised training data. Courts and regulators increasingly hold AI deployers accountable for these failures, rather than the technology vendors. Contractual structures often exacerbate this, with vendor liability caps and limited indemnities leaving deployers vulnerable.

This paper offers a practical framework to address these uninsured risks, covering:

  • Gaps in current insurance policies and why they fail to respond
  • Real-world examples of AI-driven losses and litigation
  • Emerging AI-specific insurance products and their coverage
  • Strategies for designing insurance solutions that address genuine AI failure modes

Key takeaway

AI introduces a growing category of uninsured risks. Enterprises require explicit protection for AI-driven liabilities, supported by robust governance and smarter insurance program design. This paper outlines how to achieve this.

Who should read this?

  • Underwriters, brokers and reinsurance professionals
  • Risk managers and legal teams at AI-adopting enterprises
  • C-suite leaders responsible for AI strategy and governance

What you'll learn

  • A clear understanding of AI-native risks and liability
  • Insights into current and emerging insurance solutions
  • Practical guidance on designing comprehensive AI insurance products

READ THE PAPER

Author Information