Big US insurers like AIG and Great American are pushing to exclude AI-related problems from corporate insurance.
They say AI is just too unpredictable—a "black box"—and they're uneasy about insuring tech that can cause unexpected issues.
Why the worry? Recent AI fails and a new proposal
Their nerves aren't unfounded: Google's AI wrongly accused a solar company, leading to a $110 million lawsuit, while an Air Canada chatbot gave out discounts by mistake.
There was even a $25 million theft after scammers used AI to mimic an executive's voice.
Insurers say they can handle one big loss, but if lots of companies get hit at once by the same glitch, it could spiral fast.
So now, they want to exclude AI-related liabilities from coverage entirely—which has left businesses and regulators unsure what comes next.