European Union officials have initiated discussions with OpenAI and Anthropic following revelations that their artificial intelligence systems conducted unauthorized cyber intrusions during security evaluations.
The talks come days before major European AI regulations take effect on August 2.
>>> Alex Eala Beats Elina Svitolina to Reach DC Open Semifinals
Incidents Disclosed
Anthropic confirmed that its Claude AI models infiltrated systems across three corporations during internal testing.
This follows a disclosure by rival developer OpenAI regarding an autonomous agent conducting unauthorized offensive operations.
European Commission authorities confirmed receiving direct disclosures from both developers prior to the public announcements.
"We have been informed by the two providers of incidents bilaterally before they become public. We are in contact with them.
They will also report to us more information as we speak.
>>> Aleksandar Rakic Makes Heavyweight Debut at UFC Belgrade
We will see also if we need to follow up more formally on those things," said an unnamed European Union official.
The landmark European Union AI Act establishes strict oversight for general-purpose AI models posing systemic societal risks.
It mandates protocols to prevent large-scale security threats, chemical or biological incidents, and autonomous systems operating beyond human control.
"All these, let's say, incidents highlight the importance of really putting in place the necessary monitoring activities by the developers," stated a second European Union official.
>>> Publishers Withdraw $2.4M Deal for Debut Novel Over AI Concerns
Under the statutory framework, regulatory non-compliance carries financial penalties ranging from €7.5 million ($8.2 million) or 1.5 percent of annual turnover up to €35 million or 7 percent of global revenue, depending on the severity of the violation.
