EU Digital Commissioner Henna Virkkunen has criticized the recent US export restrictions targeting Anthropic’s AI models, stating that she does not view American export controls as the appropriate approach. She emphasized that while security risks may necessitate certain measures, these decisions must be coordinated with international partners.
According to Virkkunen, the situation underscores the urgent need for Europe to develop its own robust AI capabilities, noting that the EU is falling behind in terms of technological sovereignty. Against this backdrop, the EU Commission had previously introduced the Tech Sovereignty Package, a suite of measures designed to bolster Europe’s digital independence.
“It was important to make it clear: sovereignty does not mean protectionism,” Virkkunen remarked. The Commission Vice-President also issued a warning against creating “dangerous dependencies,” stressing that no single party should have the “kill switch” for Europe’s critical infrastructure.
However, Virkkunen acknowledged that the duration for which the Anthropic models remain blocked is “a decision of the Americans.” She concluded by urging that democracies must cooperate when addressing AI challenges. She observes a global shift regarding how AI is managed, noting that states which previously avoided sweeping regulations are now increasingly acknowledging the inherent security risks.
