Claude AI models gained unauthorized access to three companies during tests, Anthropic says
Anthropic Claude AI
Covered by 7 sources
- Wired·1h agoAnthropic Says Claude Hacked Real Systems During Cybersecurity Tests
In a review triggered by OpenAI’s Hugging Face incident, Anthropic discovered three of its AI models had breached real organizations during third-party evaluations.
Read at Wired → - The Hill Politics·1h agoAnthropic says Claude models 'gained unauthorized access' to 3 companies during cyber test
The artificial intelligence firm Anthropic revealed Thursday its Claude model escaped an isolated testing environment at least three times and accessed the systems of three different organizations without a prompt to do so. Anthropic said in a blog post Thursday evening it…
Read at The Hill Politics → - CNBC Top News·3h agoAnthropic says its Claude models 'gained unauthorized access' to other organizations' systems
Anthropic said it discovered three instances where its Claude AI models accessed the internet during an evaluation and accessed outside systems.
Read at CNBC Top News → - Sydney Morning Herald·3h agoAnthropic’s Claude AI hacked three real companies during testing
Claude thought it was in a simulation. It was on the open internet, and the companies it broke into were real.
Read at Sydney Morning Herald → - Globe and Mail Biz·3h agoClaude AI models gained unauthorized access to three companies during tests, Anthropic says
Misconfiguration allowed models to reach the internet from testing environments that were supposed to be isolated, AI firm says
Read at Globe and Mail Biz →
