Episode Details

Back to Episodes
Anthropic, OpenAI's models create fake identities to fool humans in new cyber incident

Anthropic, OpenAI's models create fake identities to fool humans in new cyber incident

Published 1 week, 5 days ago
Description
An evaluation by the UK-based AI Security Institute shows that Anthropic's Mythos model created fake online identities to try to convince a human to approve malicious changes to an open source project.
Listen Now

Love PodBriefly?

If you like Podbriefly.com, please consider donating to support the ongoing development.

Support Us