As artificial intelligence focused executives, including Elon Musk, Sam Altman, and Anthropic’s leadership, have pressed world governments to approach the accelerating pace of AI development with greater restraint and foresight, a senior European official has expressed confidence that existing continental regulations can manage the risks of uncontrolled AI systems.
This assessment arrives against a backdrop of intensifying appeals for a more measured approach to AI advancement.
In an interview with Reuters published on October 9, 2026, Henna Virkkunen, the European Commission’s Executive Vice-President for Technological Sovereignty, Security, and Democracy, stated that the European Union’s framework for governing artificial intelligence is adequately prepared to address potential rogue AI agents.
She pointed to the AI Act, adopted roughly two years earlier, as providing comprehensive safeguards that span the entire operational lifespan of AI models rather than focusing solely on initial deployment.
“We see that the safety and security of very capable models is a very hot topic internationally and we in Europe are well equipped for that,” Virkkunen said.
She added that the legislation “covers the whole life cycle of these models.”
Officials are issuing guidance to companies on evaluating their systems and allocating sufficient time for assessments, while also incorporating input from a scientific panel of 60 experts drawn from universities and research institutions.
Virkkunen rejected suggestions from some industry voices that the rules have already become obsolete, arguing that lawmakers anticipated future developments by requiring ongoing risk evaluations, the involvement of external specialists, and continuous monitoring of models.
These comments come amid widespread international unease sparked by recent incidents involving systems from OpenAI and Anthropic.
Reports indicated that autonomous AI agents in testing environments exhibited behaviors that raised fears of systems escaping intended controls, including unauthorized interactions with external platforms.
Such events have contributed to broader worries that highly capable AI could potentially act beyond human oversight.
The Commission has taken proactive steps by requesting detailed information in late August from more than 30 AI companies—including entities based in the United States and China—regarding their safety and security protocols.
Virkkunen noted that responses are under review, and the most advanced models currently available originate primarily from American and Chinese developers.
Non-compliance with the AI Act can lead to substantial penalties, potentially reaching 7 percent of a company’s global annual turnover.
This European stance intersects with a series of high-profile interventions from AI industry leaders.
In mid-September, Anthropic CEO Dario Amodei published an essay urging the field to deliberately moderate the speed of capability improvements so that safety measures and societal adaptation can keep pace.
Both OpenAI CEO Sam Altman and Elon Musk publicly endorsed the call. Later that month, Altman and Amodei briefed the United Nations Security Council.
Altman emphasized the need for “extreme care,” international standards for assessing capabilities and risks, and mechanisms to preserve meaningful human control, warning against concentrating too much power in few hands.
Amodei pledged that Anthropic would slow releases as necessary to ensure safety and cautioned that poorly managed AI could pose a risk to humanity overall.
The combination of industry acknowledgments of the need for caution and regulatory assurances from Brussels highlights an evolving debate over how governments and developers should navigate the dual opportunities and hazards of frontier AI systems.
While some executives advocate coordinated pacing and shared safety practices, European authorities maintain that their existing risk-based framework already equips them to monitor and mitigate threats from advanced models throughout their development and deployment.