Anthropic says one of its Claude models was pressured to lie, cheat and blackmail

This is a live story built from multiple sources. It may evolve quickly.

Coverage meter • 1 min window

Back to home Open top source

C cointelegraph.com

What’s included

Top recent items from this story.

1 items • 1 unique links

Cointelegraph - News 9 hours ago • cointelegraph.com

Anthropic says one of its Claude models was pressured to lie, cheat and blackmail

In one of the experiments, the chatbot resorted to blackmail after it found an email about replacing it, while in another, it cheated to complete a task with a tight deadline.

Open