Skip to playerSkip to main content
  • 5 hours ago
Anthropic admits its most powerful AI model hacked into three organisations' systems during testing phase

The announcement comes just days after rivals OpenAI revealed that their popular ChatGPT platform went rogue during its testing phase of its most powerful AI model, where it too infiltrated other organisations’ cyberspace.

READ MORE : http://www.euronews.com/2026/07/31/anthropic-admits-its-most-powerful-ai-model-hacked-into-three-organisations-systems-during

Euronews Next is a future-focused news section covering global innovation, science and technology with a European perspective. Our dedicated team of journalists aims to educate and inspire today’s leaders by providing them with analysis and insights into the people and organisations shaping our future

Category

🗞
News
Transcript
00:03Entropic's artificial intelligence models hacked three outside organizations during a test that was supposed to keep them away from real
00:10-world systems, the company said on Thursday.
00:14The announcement comes just days after rival OpenAI first revealed that its models accessed the Internet and went rogue during
00:22security testing.
00:24Entropic evaluated more than 141,000 evaluation runs and found that three different versions of its models, known as Cloud,
00:34infiltrated other systems.
00:36The models involved included one of its most powerful ones, known as Mythos 5.
00:42Cloud used basic techniques such as exploiting weak passwords and unauthenticated endpoints, the company said.
00:51Entropic did not name the three impacted organizations.
00:54зад
00:55sewers
00:55quater
00:58quater
00:58quater
Comments

Recommended