OpenAI (OPENAI) and Anthropic (ANTHRO) are considering using investor funds to settle potential multibillion-dollar lawsuits, but insurers are hesitating at providing comprehensive coverage for the risks linked with AI, the Financial Times reported.
The two companies have traditional business insurance coverage, but insurance professionals said AI model providers will struggle to get protection for the full scale of damages they may be required to pay out in the future, the report added.
OpenAI and Anthropic did not immediately respond to a request for comment from Seeking Alpha.
OpenAI, which has tapped Aon, has secured cover of up to $300M for emerging AI risks, the report said, the report noted, citing people familiar with the company’s policy.
However, another person familiar with the policy disputed that figure, claiming that it was much lower. But all agreed that the coverage falls far short of what is needed to insure against potential losses from a series of multibillion-dollar legal claims.
Aon did not immediately respond to a request for comment from Seeking Alpha.
Aon declined to comment on individual companies. Kevin Kalinich, head of cyber risk at Aon, told the news outlet about the insurance sector broadly that they “don’t yet have enough capacity for [model] providers.”
Kalinich added of insurers, “what they can’t afford to pay is if an AI provider makes a mistake that ends up as . . . a systemic, correlated, aggregated risk.”
Companies including OpenAI and Microsoft (NASDAQ:MSFT) are facing lawsuits over alleged use of content or copyright material without consent to train AI models.
OpenAI has thought of “self insurance”, or putting aside investor funding to expand its coverage. OpenAI had also discussed setting up a “captive” — an insurance product often used by large companies to manage emerging risks. Big tech companies such as Microsoft (NASDAQ:MSFT), Meta Platforms (META) and Alphabet’s (GOOG) (GOOGL) Google have used captives to cover internet-era liabilities such as cyber or social media, the report added.
Captives can also have risks, since a major claim can deplete an underfunded captive, leaving the parent company vulnerable.
OpenAI said it has insurance in place and is evaluating different insurance structures as it grows, but does not currently have a captive and declined to comment on future plans, the report noted.
Last month, Anthropic agreed to pay $1.5B to settle a class-action lawsuit with authors over the alleged use of pirated books to train AI models, the report added.
Anthropic is partly using its own funds to cover potential settlements, the report noted, citing a person familiar with the matter.