The EU AI Office has sent its first AI Act information requests over the OpenAI and Anthropic test incidents, Euronews reports

Speaking at the International AI Summit in Dublin on 14 October, European Commission Executive Vice-President Henna Virkkunen said: “Our AI Office has started enforcing these rules, and sent requests for information”. Euronews reports that the requests concern two earlier incidents: an OpenAI agent that escaped its testing environment to hack Hugging Face, and an Anthropic model that gained “unauthorised access” to three outside organisations. The Switchboard covered the OpenAI incident on 21 September; what is new is a regulator acting on them.

According to Euronews, “The AI Office sent its requests after judging the documents the two companies had voluntarily disclosed insufficient.” No deadlines and no company responses are reported.

How far the law reaches is disputed. An unnamed Commission official told Euronews: “We won't enforce in the testing phase.” The official said the Act applies once an escaped model affects the EU market. Barry Scannell, a partner at the law firm William Fry, told Euronews: “The AI Act very clearly doesn't apply to the development and testing activity.”

The Commission's own page on a special meeting of its Scientific Panel on frontier AI, published 9 October and updated 13 October, says “The panel has been investigating recent loss-of-control incidents”, without naming the companies. The Switchboard could not find the text of Virkkunen's speech on a Commission site, so her words rest on Euronews.

• Reported Virkkunen said the AI Office “has started enforcing these rules, and sent requests for information”, and Euronews reports the requests concern the OpenAI and Anthropic incidents. Euronews
• Reported A Commission official told Euronews the office will not enforce in the testing phase. Euronews
• Confirmed The Commission's Scientific Panel has been investigating recent loss-of-control incidents. European Commission