Recent AI infringement lawsuits, particularly those involving copyright claims, are already shaping the direction of AI insurance policies in several significant ways:
Expansion and Specialization of Coverage
- Rise of AI-Specific Liability Policies: The surge in lawsuits, such as Thomson Reuters v. Ross Intelligence and actions against Meta for using copyrighted materials in AI training, is driving demand for specialist insurance products. These policies are designed to address unique AI risks, including copyright infringement, algorithmic bias, privacy breaches, and performance failures(sources: Costero Brokers, IPWatchdog).
- Broader Intellectual Property (IP) Coverage: Standard Commercial General Liability (CGL) policies often provide limited protection for copyright infringement related to AI outputs, typically only covering "advertising injury." Insurers are responding by offering more comprehensive IP liability or media liability policies tailored to AI-related exposures, ensuring coverage for lawsuits over unauthorized use of data in AI training and content generation (source: IPWatchdog).
Policy Exclusions and Underwriting Adjustments
- Emergence of AI-Related Exclusions: Insurers, wary of the unpredictable and potentially massive liabilities from generative AI, are increasingly introducing exclusions in media liability, cyber-risk, and technology errors & omissions (E&O) policies. These exclusions may specifically target content created or decisions made by AI, narrowing the scope of coverage for certain AI-related claims.
- Underwriting Scrutiny: Insurers are asking more detailed questions about how companies use AI, how their models are trained, and what data sources are used. Inaccurate or incomplete disclosures about AI use could lead to policy rescission or denial of claims (source: Stoel Rives).
Financial and Legal Risk Mitigation
- Coverage for Legal Defense and Settlements: With courts signaling a willingness to hold AI companies liable for copyright infringement (as in the Thomson Reuters v. Ross decision), insurance policies are being structured to cover not just damages, but also the substantial legal expenses, settlements, and judgments that can arise from such claims.
- Emphasis on Licensing and Compliance: Insurers may require proof of proper licensing for training data and robust risk management practices as a condition of coverage, to minimize exposure to infringement claims (source: IPWatchdog).
New and Evolving Products
- AI-Specific Endorsements: Some insurers are introducing endorsements that specifically cover "Machine Learning Wrongful Acts" (e.g., unauthorized use of data for AI training) or "Data Poisoning Wrongful Acts" (e.g., introduction of false data into training sets) (source: Stoel Rives).
- Performance Failure and Hallucination Coverage: Products like Munich Re’s aiSure and aiSelf offer coverage for financial losses stemming from AI underperformance or output errors, reflecting the broader spectrum of risks associated with generative AI (source: S&P Global).
Conclusion
Recent AI infringement lawsuits are prompting insurers to both broaden and clarify coverage for AI-related risks, while also introducing new exclusions and underwriting requirements. As legal precedents develop, businesses deploying AI must review their insurance portfolios to ensure adequate protection against the evolving landscape of IP, liability, and operational risks associated with AI.
References:
- Exploring AI liability insurance: New protections for fast-evolving business risks - Costero Brokers
- Takeaways from Thomson Reuters v. Ross Intelligence: AI Litigation Risks and Insurance Solutions
- Insurers brace for claims from generative AI surge | S&P Global
- AI and Insurance – The Awkward Early Days | Stoel Rives LLP
- Make Sure You’re Covered for the AI Copyright Fight: Insurance Safeguards After Thomson Reuters v. Ross
- Law firm AI: insurance considerations | Lockton
- Artificial Intelligence and Insurance - Part II, The Story Unfolds
- AI – unintended insurance impacts and lessons from “silent cyber” | Swiss Re
- AI Blind Spot: Is the Board Exposing the Company to Uninsured AI Risks? | Law Bulletins
- Insurance coverage issues, artificial intelligence and deepfakes | Reuters
