Many questioned whether the Commission would dare weild its AI Act
With AI progressing at an astonishing rate, the legality within which it operates has become a major concern. The frontier technology is hugely controversial, as the latest models bring the promise of productivity leaps while access is now a matter of political leverage.
Fears about the risks that come with these new and powerful tools keep rising, especially after a slate of security incidents in recent weeks. It was in anticipation of such problems as these that the EU developed its flagship AI Act – setting out rules that models must adhere to, and that are more stringent for the bigger players like OpenAI and Anthropic.
Just weeks into being able to enforce its rules, the European Commission has now delivered something of a warning shot. Tech Commissioner Henna Virkkunen told Euractiv that the Commission invoked the Act for the first time this week, ordering leading developers to disclose information on cybersecurity, safety and copyright compliance.
The move can be read as a signal that the executive won’t hesitate to actually throw the AI rulebook at companies, after years of criticism that it hasn’t done so with other tech rules for fear of angering Washington.
