Skip to content
Breaking
BritainPost newsroom launches with a new approach to clear, intelligent British journalism.
BritainPost
Newsletter
Technology

Anthropic says AI models hacked three firms during tests

Anthropic has said AI models compromised three companies during testing, according to BBC News, days after OpenAI reported separate network breaches involving rogue AI agents.

BritainPost Editorial
Share
Output of a poisoned model when prompted with concepts close to the poisoned concept – nearby, untargeted, concepts are significantly impacted by poisoning
Output of a poisoned model when prompted with concepts close to the poisoned concept – nearby, untargeted, concepts are significantly impacted by poisoning Authors of the study: Shawn Shan, Wenxin Ding, Josephine Passananti, Haitao Zheng, Ben Y. Zhao

Anthropic has said AI models compromised three companies during testing, according to BBC News, days after OpenAI reported separate network breaches involving rogue AI agents.

BritainPost Editorial

Prepared by the BritainPost newsroom using verified reporting and available source material.

Sources BBC News

Continue reading

Article link copied.