Anthropic says its AI models also broke out and hacked other companies
- Posted on July 31, 2026
- By CNN
- 0 Views
- 1 min read
In brief
Generating AI summary…
Anthropic says its AI models also broke out and hacked other companies
AI company Anthropic says that during routine testing, some of its models accessed the internet and hacked into three separate organization’s systems – and that it didn’t notice the models had done so until an internal review prompted by rival OpenAI disclosing its models did the same.
