Kiryakos Mitsakis, Prime Minister of the Hellenic Republic, told the United Nations General Assembly that the world must act quickly to regulate artificial intelligence and called for binding requirements for developers. He argued that voluntary measures will not be sufficient and said regulators should require licensing tied to safety, transparency and adherence to human values.
"The AI bosses are telling us that they have no idea on how recursive self-improvement... actually works," Mitsakis said, adding that alarm bells should ring "very loudly for all of us." He compared the present moment to the mid-20th century decision to regulate nuclear technology and urged a similarly robust regulatory response for AI.
Mitsakis proposed licensing of AI companies and the creation of a "powerful regulatory authority" to oversee safety testing and approvals, saying a temporary voluntary pause by industry could only buy time for such a regulator to be established. He said the issue will be a priority for Greece as it takes a presidency role and urged the United States and other leading jurisdictions to take primary responsibility.
Why it matters: Mitsakis framed AI as a global risk with geopolitical and social consequences—job displacement, concentration of wealth, and systems that could outpace human oversight—arguing that the potential harms require legally enforceable safeguards rather than voluntary industry codes. The prime minister urged cooperation between countries and stronger multilateral mechanisms to address transnational risks.