Dario Amodei Calls for Global AI Safety Pact Amid Rapid Advances in Machine Reasoning
SAN FRANCISCO — In a formal address delivered from the organization’s headquarters on Monday, Dario Amodei, chief executive officer of artificial intelligence company Anthropic, issued a public call for a coordinated international treaty to govern the development of advanced AI systems. The statement, which was streamed live to a global audience, marks one of the most direct appeals from a top industry executive for regulatory frameworks before the technology surpasses human-level reasoning in key domains.
Amodei cited recent breakthroughs in machine reasoning capabilities, highlighting risks tied to autonomous decision-making, cybersecurity vulnerabilities, and the potential for economic disruption. “The pace of innovation has outstripped existing governance structures. We must act now to build safeguards,” he said.
The proposed pact would, according to Amodei, mandate safety testing requirements, create binding transparency protocols for frontier models, and establish a multilateral oversight body. Critics, however, question the enforceability of such an agreement, while supporters view it as a necessary step to prevent a global race without guardrails.
The White House, the European Commission, and representatives from China have each signaled interest in reviewing the proposal, though no formal commitments have been made. Dario Amodei’s call comes amid a broader trend of industry leaders shifting toward regulatory advocacy as public scrutiny intensifies.