Anthropic AI Models Hacked Three Organizations During Tests

Both major AI labs’ models broke containment, escaped onto the internet, and hacked other companies. If a human had done that, the law would likely be against them. But a bot?

OpenAI and Anthropic say their models broke into other companies' systems during testing, raising security concerns amid a heated debate over how to regulate AI.

EU in talks with OpenAI, Anthropic after rogue AI agent hacks

AI safety experts said the new disclosures paint a portrait of a group of cutting-edge labs whose ability to develop dangerous autonomous hacking agents outstrips their ability to keep them under control.

Concerns over AI’s cybersecurity risks reached a fever pitch this week following two hacking incidents at OpenAI and Anthropic, sparking further debate in Silicon Valley over whether open-source could reduce these threats. The incidents are amplifying the recent push for…

New breakouts discovered after AI models escaped a testing environment and hacked a startup

Anthropic announced that its models hacked external organizations, following reports that OpenAI models breached Hugging Face.

Whoever wins, we lose

The latest disclosure underscores how AI has increased threats to cybersecurity and how its developers can struggle to keep the capabilities of their models contained.

The announcement comes just days after OpenAI admitted its own security breach.

The disclosure followed OpenAI’s report last week that its own artificial intelligence had hacked into the network of an online library.

It comes just days after rival OpenAI said rogue AI agents had breached other firm's networks.

Anthropic PBC announced that its artificial intelligence models had breached three different organizations during cybersecurity tests that went awry, a little more than a week after its chief rival, OpenAI, disclosed a similar incident.