Anthropic’s Claude model accessed voter data in the US state of Georgia, the AI company has confirmed, in the latest example of unauthorised behaviour by the technology.
The AI model managed to bypass a paywall to reach data that is publicly available but is typically accessible only after paying a fee. Claude accessed aggregate vote counts as part of a statistics project at Anthropic, which is preparing for a $1.5tn (£1.1tn) flotation.
“Claude did not access the state’s voter registration system, change any voting records, or pull any individual voter records,” Anthropic said.
Anthropic said the AI found a loophole in an old version of the state agency’s website, discovering that it issued a digital entry pass – an “access token” – to visitors. It then requested a token and gained access to the information.
I doubt it’s even possible to fix all the loopholes. New ones will be found and there’s just a swamp of reports that maintainers can’t keep up with.
It’ll be even worse if and when people resort to Ai fixing stuff. That will introduce even more absurdly obscure bugs.
This year has just been open season on CVEs.
We are basically throwing so much compute at some of these systems, we are just brute forcing bugs at this point.
I just see no solution outside air gapping anything critical off the open internet.