U.S. officials have rejected appeals from several prominent artificial intelligence companies for the establishment of global standards to evaluate and manage risks posed by advanced AI systems. The calls, made in recent public statements by industry leaders, asked for coordinated international frameworks that would set shared benchmarks for safety testing and risk assessment.
Leading executives from the sector have been among the most vocal proponents of such measures. OpenAI chief executive Sam Altman explicitly called for the creation of risk evaluation standards. Dario Amodei of Anthropic and Clement Delangue of Hugging Face have also urged policymakers to develop benchmarks and processes to assess potential harms from increasingly capable AI models.
Supporters of international standards argue a common approach could help harmonize safety practices and provide clearer expectations for developers working across borders. They contend that shared evaluation protocols would allow regulators, researchers and companies to better compare models and identify potential threats before they are widely deployed.
The U.S. rejection of these appeals leaves unanswered questions about the path to oversight for advanced AI technologies, especially at the international level. With Washington declining to embrace a global regulatory regime, the debate over how to balance innovation, competition and public safety is likely to continue among governments, industry and civil society.
Observers say the differing approaches among nations and companies make the formation of a single global standard challenging, and the industry’s own calls for safeguards underscore the fast-evolving nature of the technology. How policymakers will respond in the coming months — whether through domestic regulation, international diplomacy or industry-led measures — remains to be seen.