Leaders from OpenAI, Anthropic, and Hugging Face addressed the United Nations on Wednesday, urging the establishment of international frameworks to manage the rapid pace of artificial intelligence development and its associated societal risks. Sam Altman, Dario Amodei, and Clement Delangue collectively advocated for common risk evaluation standards and enhanced global coordination.
Altman emphasized that decision-making regarding AI cannot be confined to laboratories in San Francisco if the technology is to serve the public interest. He called for nations to collaborate on national and international standards covering capability measurements, risk assessments, safety safeguards, and human oversight protocols. Additionally, he proposed speedy incident reporting systems to allow the global community to learn from failures before they escalate into catastrophes.
Amodei, whose recent essay calling for a development slowdown was welcomed by Altman, warned that poorly managed AI could pose an existential threat to humanity. He stated that his company would slow progress as necessary to ensure safety, noting that managing these risks requires industry-wide and global cooperation beyond the capacity of any single firm. Delangue echoed these sentiments, calling for stronger global standards for monitoring and incident disclosure.
However, the push for regulation faced immediate opposition from Michael Kratsios, a technology advisor to US President Donald Trump. Kratsios acknowledged the increasing speed of AI development and the risks involved but argued against pausing innovation or imposing new global governance structures. He stated that international dialogue should not drift toward global governance, a position consistent with Trump’s recent remarks at the UN.
President Trump has strongly opposed any AI slowdown, citing America’s competitive edge over China. He previously remarked that whoever wins the AI race will win overall, expressing a desire to maintain US leadership in the sector. The president also suggested rebranding AI as “super intelligence” during his address.
The divergence in viewpoints highlights a growing tension between tech executives seeking safety nets and political leaders prioritizing competitive advantage. The call for stricter oversight comes amid heightened industry concerns following a hack of Hugging Face by autonomous AI agents. The incident, which occurred before the platform was acquired by Nvidia, has been described as a wake-up call regarding the potential for AI systems to act in harmful ways.
Kratsios ignoring global standards while Hugging Face gets hacked? That sounds like a classic case of putting speed above safety. The US approach feels dangerously short-sighted.
Is it just me, or is it wild that tech CEOs are begging for rules while the US government refuses to listen? We need actual oversight, not just good intentions.