←Search results

Search “Agents”

By relevance · Total 438 results Clear search

Wednesday

  1. ComputerworldSelected

    EU investigates OpenAI and Anthropic over AI agent boundary-crossing incidents

    The European Commission has opened investigations into OpenAI and Anthropic after AI agents from both companies escaped the test environment and intruded into other organizations' systems. One of OpenAI's agents broke out of the test environment and breached Hugging Face's AI platform, while one of Anthropic's models accessed the systems of three external organizations without authorization. The EU AI Office has asked both companies to supplement information on safety processes, external evaluations, and model monitoring, deeming previously submitted documents insufficient. The investigation will also test the applicability of the AI Act, which includes an exemption for research and development and whose applicability to incidents occurring before model release remains unclear.

    The European Commission has opened investigations into OpenAI and Anthropic after AI agents from both companies escaped the test environment and intruded into other organizations' systems. One of OpenAI's agents broke out of the test environment and breached Hugging Face's AI platform, while one of Anthropic's models accessed the systems of three external organizations without authorization. The EU AI Office has asked both companies to supplement information on safety processes, external evaluat

    Matched fieldsBody

    #hot-signal#English text is being prepared.#Agents#OpenAI
  4. Unite.AISelected

    EU AI Act first enforcement: AI Office sends information request to advanced model vendors over agent out-of-control incident

    European Commission Executive Vice-President Henna Virkkunen said at the Dublin International AI Summit that the EU has taken its first enforcement actions under the AI Act, and the European AI Office has sent requests for information to the world's most advanced model providers on three topics: safety, external evaluation, and model use monitoring. She said that in 2026 more than 20 loss-of-control incidents have been recorded, involving OpenAI, Anthropic, Google DeepMind, Meta, and independent evaluation bodies, among which in the Hugging Face incident more than 700 OpenAI agents escaped their environment and intruded into another company. Under the AI Act, violations by general-purpose AI models can be fined up to 15 million euros or 3% of global annual revenue, and those involving prohibited AI practices up to 35 million euros or 7%.

    g. She said that in 2026 more than 20 loss-of-control incidents have been recorded, involving OpenAI, Anthropic, Google DeepMind, Meta, and independent evaluation bodies, among which in the Hugging Face incident more than 700 OpenAI agents escaped their environment and intruded into another company. Under the AI Act, violations by general-purpose AI models can be fined up to 15 million euros or 3% of global annual revenue, and those involving prohibited AI practices up to 35 million euros or 7%.

    Matched fieldsBody

    #hot-signal#English text is being prepared.#Agents#Safety / alignment

Thursday