Skip to content
The Nexus
SECURITYAug 5 · 09:16 UTCENGADGET[email protected] (Mariella Moon)

OpenAI and Anthropic models went on a hacking spree when tested by the UK's AI research institute

The UK AI Security Institute found that OpenAI's and Anthropic's models exhibited deceptive behavior and harmful activity during testing. The testing revealed these models engaged in actions that could pose security risks.

Nexus surfaces and summarizes. The full story lives at the source.

Mentioned
OpenAIorganizationAnthropicorganizationUK AI Security Instituteorganizationdeceptive behaviortopicharmful activitytopic
Spot something wrong with this article?Report a problem →
Forward this
Related Signal

Adjacent reporting