AI
Our Brand - Insurance for the AI Age
Generative AI lawsuits provide an insight as to what is driving real word AI risk. Testudo's data provides a map of understanding for liability in the AI economy to date. Reach out to learn more!
Federal and state generative AI-related lawsuits are one type of high quality signal for real world AI risk. The number of GenAI filings increased 137% in 2025, an alarming signal that litigation is rising exponentially as AI usage expands, according to a new report released today by Testudo. As AI deployments proliferate across enterprises in the U.S. and insurance coverage is pulled back by insurers, “when it comes to AI liability risk, putting your head in the sand and hoping it goes away isn’t a plan” - George Lewin-Smith.
The State of Play is from the dedicated research lab of Testudo Global, Inc., where we use data insights from real world AI risk underwriting and pricing to provide market commentary. Email us to request it!
Key Findings
- Generative AI lawsuit filings are up 137% in 2025 and rising
- As AI usage expands across the U.S. economy, litigation is increasing exponentially.
- Model hallucinations define only 4.9% of generative AI lawsuits.
- A very small number of U.S. generative AI lawsuits involve claims related to AI ‘hallucinations’, which are instances where AI systems generate false or misleading information that causes harm. This highlights that the performance of an AI model has limited impact on a firm's AI liability exposure.
- 32% of generative AI lawsuits are class actions
- One third of generative AI lawsuits are class actions. Class actions represent a particularly acute source of aggregate exposure for defendants, as they concentrate large numbers of similarly situated claimants into a single proceeding, amplifying damages, defense costs, and settlement pressure.
- $5M is the median demand amount for an AI lawsuit
- While 70% of cases seek less than $10M, approximately 16% involve demands exceeding $100M.
- There is an extreme loss concentration for AI lawsuits
- The top 5% of cases account for 99% of total demanded amounts. Of course, demands are not settlements.
- Over 32% of publicly traded U.S. companies are using AI
- Generative AI liability risk is real, increasing, causing damages and sitting silently on the balance sheets of enterprises that passively deploy generative AI.
- AI deployment rates and litigation risk by industry
- Below is an overview of AI deployment rates per industry (NAICS codes) and a view of AI litigation risk per industry based on our data.
- Bonus: 2026 is the year of the Agent
- AI agents materially increase risk exposure because they can act autonomously and execute at scale. While human-in-the-loop review and other braking mechanisms can reduce the likelihood of serious harm, they constrain system speed and autonomy, undermining the efficiency gains that drive adoption.