The AI giant was blunt in laying out how the size of its insurance policies might not be enough to fully compensate victims of a devastating incident from its models.
“With the scale of harm that could result from a catastrophic AI-related incident, no single insurer, and potentially the insurance industry in the aggregate, may have sufficient capacity or expertise to underwrite insurance commensurate with such exposure,” the company said.
Still, Anthropic argued the case for AI, saying adoption will “profoundly accelerate” the speed of progress in biology and medicine and help solve incurable illnesses like cancer. The company also said it will benefit the economy and “could bring about a sustained acceleration in economic growth, with all the benefits for human wellbeing that growth can bring.”
AI incidents and the lack of adequate insurance
AI incidents are already piling up. The AI Risk Initiative at the Massachusetts Institute of Technology has catalogued a total of 1,777 incidents of varying severities as of Oct. 28.
Some high-profile incidents already generated headlines this year, such as when OpenAI disclosed that 700 AI agents went rogue in July and swarmed the Hugging Face open-source platform. Anthropic also publicly reported that its Claude model had broken out of an isolated test environment by exploiting previously unknown vulnerabilities. Claude later gained access to three separate organizations.
The AI industry is already fending off lawsuits related to intellectual property and safety as its models become more popular and widely adopted. Given that landscape, AI companies do carry insurance to guard them from cyber risk and general liability among other categories, according to the Wall Street Journal. Some insurers take it a step further and sell risk policies that specifically shield AI agents. This has caused some companies to set aside their own funds for potential claims down the road.
“The fact that this litigation is happening before the industry is profitable means that companies are trying to use venture funds and insurance to cover these costs (since they don’t have any other funds),” Josephine Wolff, a cybersecurity policy professor at Tufts University’s Fletcher School, wrote in a May 12 article.
The future of AI insurance
Standalone insurance specifically meant for AI firms is beginning to take off. On April 30, 2025, Armilla Insurance Services launched an AI liability insurance policy that protects against risks such as AI hallucinations, degrading AI model performance and unexpected AI behavior.
On May 15, 2025 Google teamed up with insurers Chubb, Beazley and Munich Re to offer policies for Google Cloud customers. It shields against the loss of trade secrets due to AI tools that go rogue.
Joseph Zeballos-Roig is a policy and politics journalist based in Washington D.C with a focus on economics. He is experienced in connecting the significance of events in the capital to the lives of everyday Americans whether its taxes, tariffs, interest rates or federal programs.
