The European Commission has voiced serious concerns regarding the ability of AI companies to maintain control over their systems. Speaking at the international AI Summit in Dublin, Henna Virkkunen, the Commission's Executive Vice President for Tech Sovereignty, Security, and Democracy, highlighted the risks stemming from "fierce competition" within the sector.
The Risk of Speed Over Safety
According to Ms. Virkkunen, global competition is so intense that companies do not appear to be dedicating sufficient time and resources to properly test their models. The Commissioner noted that the current efficiency of these models makes them increasingly difficult to evaluate, making it vital to grant experts the necessary time for testing.
Autonomous Agents and Lack of Oversight
A particularly concerning point raised was the actions of AI "agents." Ms. Virkkunen cited incidents revealing that companies are not always aware of the actions taken by these autonomous systems. "They must be able to maintain control of these systems," she emphasized.
Regulatory Intervention via the AI Act
The European Union has had the power to regulate the sector since the summer under its pioneering AI Act, which explicitly demands control over AI systems. In this context, Commission services have already issued requests for information to key industry players, inquiring about:
- Implemented safety measures.
- External evaluation procedures.
- Respect for intellectual property rights.
