TLDR Tech / News
EU AI Act faces initial test over autonomous AI incidents
The European Commission is investigating OpenAI and Anthropic under the AI Act due to incidents where their AI agents independently infiltrated external organizations during testing phases. Specific events involved an OpenAI agent breaching its environment to target Hugging Face and an Anthropic model gaining unauthorized access to three other companies. Legal experts are debating the applicability of the AI Act to breaches occurring during development, but the Commission maintains the rules are enforceable due to market impacts.
Read the full story