Anthropic says the “sinister” portrayal of the AI ​​was responsible for Claude’s blackmail attempts

Fictional depictions of AI could have a real-world impact on AI models, according to Anthropic.

Last year, the company said that during pre-release tests involving a fictional company, Cloud Opus 4 often tried to blackmail engineers into avoiding replacing it with another system. Anthropic later published research suggesting that models from other companies had similar issues with “misalignment.”

Anthropic has apparently done more work on this behavior, claiming in a post on

The company went into more detail in a blog post stating that since version 4.5 of Claude Haiku, Anthropic models “never participate in extortion.” [during testing]where previous models sometimes did this up to 96% of the time.

What explains the difference? The company said it found that training on “documentation about Cloud’s architecture and fictional stories about how the AI ​​behaves impressively improves compliance.”

Related, Anthropic said it has found that training is most effective when it includes “the principles behind biased behavior” and not just “the display of biased behavior alone.”

“Doing both appears to be the most effective strategy,” the company said.