European AI Office Requests Data from OpenAI and Anthropic
The European AI Office sent its first official requests for information to OpenAI and Anthropic, covering model security, evaluation, and post-deployment monitoring.
The European AI Office sent its first official requests for information to OpenAI and Anthropic, according to an announcement by European Commission Executive Vice-President Henna Virkkunen at the International AI Summit in Dublin.
The requests cover model security, evaluation by external experts, and monitoring behavior in real-world use after release to users.
The action follows incidents last summer involving an agent from OpenAI and a model from Anthropic, and the step is an investigative procedure, not a fine.
Fines for violations involving general-purpose models could reach up to 15 million euros or 3% of global turnover, according to Luwai platform.
What do these terms mean?
Request for Information: An official letter issued by a regulatory authority to obtain data or clarifications, serving as a preliminary step prior to imposing fines.
General-Purpose Models: Powerful artificial intelligence models that can be used across many applications and are not dedicated to a single task.
External Evaluation: Assessment of the model by an entity independent of the developer company to verify its safety and performance.
Post-deployment Monitoring: Tracking model behavior after making it available to users to ensure continued compliance with rules.
Weekly Newsletter
Read between the lines before everyone else. Decode the most important economic, tech, and decision-maker movements in the region.. in 5 minutes every Saturday.
