Insurance companies are increasingly excluding coverage for artificial intelligence-related harms, with major insurers such as W. R. Berkley and Verisk/ISO adding exclusions to their policies. The RAND Corporation's report highlights a significant mismatch between the rapid deployment of AI by businesses seeking to protect themselves from liability and the insurance industry's reluctance to underwrite these novel risks. This shift could substantially slow corporate AI adoption if coverage becomes unavailable, forcing companies to delay projects to meet their fiduciary obligations.

The report documents widespread AI-related incidents and legal challenges: the Artificial Intelligence Incident Database records 713 documented incidents spanning multiple categories, including misinformation and manipulation (586 cases), deepfakes and synthetic media (346), and hallucination or factual errors (215). Additionally, approximately 250 US lawsuits related to AI are ongoing, largely concerning copyright and intellectual property disputes, though also covering privacy, fraud, discrimination, and negligence. Multiple US states have introduced laws addressing AI-generated content, automated decision-making, and political advertising, creating additional regulatory uncertainty for businesses deploying AI technology. Despite these obstacles, companies continue investing in and deploying AI systems.