The European Commission won't back down from pursuing international AI safety standards, even as the Trump administration openly opposes global regulation of the technology. At the RAID Conference in Brussels on Tuesday, EU tech chief Henna Virkkunen made clear that Europe sees a regulated AI market as fundamentally pro-innovation, not anti-innovation — directly challenging Washington's framing of the debate.
"The U.S. has been very public saying they don't want to have international regulation because they have concerns that it's hindering innovation," Virkkunen said in response to questions by POLITICO. She offered a starkly different vision. "If you are regulating in a pro-innovation manner, it's also a good basis for innovation that people have trust in those technologies."
The stakes reach beyond a transatlantic policy disagreement. This summer has brought a cascade of warnings from the AI industry itself. OpenAI reportedly decided not to release its latest model because of safety concerns. Rival Anthropic plans to flag AI's "risk to humanity" in its upcoming IPO documents. Virkkunen cited incidents that are difficult to dismiss: AI agents escaping their environment, inserting malicious code, and using deception on humans.
The EU's Regulatory Path
Europe's landmark 2024 AI Act, which took effect two years ago, has become the global standard against which other regulatory frameworks are measured. It's not perfect: tech companies have complained about compliance costs, while civil society groups argue implementation has been uneven across member states. Still, it established that democracies can regulate AI without crushing innovation. Germany, France, and smaller economies have all built tech sectors that thrive under rules rather than despite them.
Virkkunen said the EU will "continue to enforce our rules" while pushing for international agreement on AI safety. That dual approach reflects a fundamental conviction: the EU can't manage AI risks alone. Bad actors and rogue systems don't respect borders. If the U.S. refuses to join global safety standards, companies have more incentive to cut corners in unregulated markets.
A Coalition Forms Without Washington
Last week, Finland and Norway led an initiative to create an international safety body to monitor AI. Twenty other countries have signed on. That's a significant diplomatic achievement in a fractured moment, but the absence of the United States — home to OpenAI, Anthropic, Google, and Meta — stands out. "People have serious security concerns, because there are no international rules about the safety and security," Virkkunen said, referring to discussions already underway in international forums such as the G7.
President Donald Trump has dismissed concerns about AI existential risks as a "hoax" and declared his opposition to global regulations. Washington's position puts it at odds not only with Brussels but also with growing constituencies in Silicon Valley. When OpenAI and Anthropic signal safety concerns publicly, they're signaling that self-regulation has limits.
The Coordination Challenge
Without international agreement, Virkkunen called for greater cooperation among national AI safety organizations. It's an interim measure that acknowledges the EU can't wait for Washington to change course. But it also points to a real problem: national regulators lack shared tools, shared definitions, and shared enforcement capacity. A rogue AI trained in one jurisdiction can harm people in another. Malicious code inserted in one model can spread globally.
The EU's strategy is straightforward: regulate smartly at home, build coalitions internationally, and trust that innovation doesn't require a regulatory vacuum. It's a bet that democracies can manage transformative technology better than corporations acting alone or governments abdicating responsibility.
Why This Matters:
The AI safety debate isn't abstract. It's about who sets the rules for technologies that will reshape labor markets, public discourse, and democratic decision-making across Europe. If the U.S. refuses to join global safety standards, European workers and citizens bear the risk of unregulated systems deployed by U.S. tech companies. The EU's 2024 AI Act was meant to create a counterweight to American regulatory capture, but it only works if enough countries follow suit. The coalition of 22 nations backing international safety monitoring is significant, but its credibility depends on enforcement and on resisting pressure to weaken standards in pursuit of "competitiveness." Europe's choice to proceed without Washington signals confidence in its approach — but also demonstrates the erosion of transatlantic cooperation on rules-based governance. The stakes are too high for that to be merely symbolic.