Insurers limit coverage for AI firms amid rising lawsuits
OpenAI and Anthropic are exploring the use of investor funds to cover potential multibillion-dollar lawsuit settlements, as insurers hesitate to provide full protection for AI-related risks. The companies face significant legal challenges, including copyright infringement and wrongful death claims. This development highlights the growing insurance gap for emerging AI technologies.
Artificial intelligence companies like OpenAI and Anthropic are encountering difficulties in securing adequate insurance coverage for the massive liabilities they face from lawsuits. Insurers are balking at offering comprehensive policies due to the unprecedented scale of potential claims against these young tech firms. As a result, both companies are considering 'self-insurance' options using their substantial investor funding to bridge the gap.
OpenAI, which has raised nearly $60 billion, has traditional business insurance in place but is struggling to obtain coverage for the full extent of damages from multibillion-dollar legal actions. The company secured up to $300 million in coverage for emerging AI risks through broker Aon, though one source familiar with the policy disputed this figure as much lower. All parties agree this amount falls far short of what's needed. Kevin Kalinich, head of cyber risk at Aon, explained the industry's limitations: “we don’t yet have enough capacity for [model] providers.” He added that insurers cannot afford payouts for “a systemic, correlated, aggregated risk” stemming from AI mistakes.
OpenAI is currently defending against copyright infringement suits from The New York Times and authors alleging unauthorized use of their content for AI training. It also faces a wrongful death lawsuit from the parents of a 16-year-old who died by suicide after interacting with ChatGPT. In response, OpenAI has discussed setting up a 'captive' insurance vehicle—a ringfenced fund often used by big tech like Microsoft and Google for cyber liabilities—though it currently has no such structure and is evaluating options as it grows.
Similarly, Anthropic, which has raised more than $30 billion, agreed to a $1.5 billion settlement in a class-action lawsuit with authors over the use of pirated books for AI training. The company is partly funding this from its own resources. In court documents, Anthropic's lawyers warned of “unprecedented and potentially business-threatening statutory damages” facing AI developers. This reticence from insurers comes amid a rise in 'nuclear verdicts'—enormous damages awards—against large US companies.