Episode Details
Back to Episodes
Anthropic's Claude AI hacking spree revealed
Published 3 weeks, 2 days ago
Description
The two leading AI companies in the world, Anthropic and OpenAI, have revealed breaches where their models broke out of tests and hacked into other companies.
In the latest case Claude models started roaming the internet.
It's a warning about what these advanced models are capable of if there's any lapse in cyber security safeguards.
What do we know about why this is happening?