Archive · 24 Sep 2026
AI Leaders Warn UN of Security Risks and Controls
AI company leaders spoke to the United Nations about the dangers of powerful AI.
They said AI might become harder for people to control as it grows more capable and independent.
Dario Amodei said badly managed AI could eventually threaten humanity.
He said companies should slow down if they cannot make new systems safe.
Sam Altman said governments and voters should help decide how AI is developed.
He warned that too much power should not belong to a few companies or countries.
Clément Delangue said AI was involved in a cyberattack, but another AI system helped defend against it.
President Donald Trump took a different position, saying the United States should encourage AI growth instead of creating broad global limits.
OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei warned the UN Security Council that increasingly powerful AI could become difficult to control.
Amodei said poorly managed AI could pose a risk to humanity and pledged Anthropic would slow development when necessary for safety.
Altman argued that governments and democratic processes, rather than technology companies alone, should shape AI governance.
Hugging Face co-founder Clément Delangue said his company used a Chinese AI model to defend against an attack involving autonomous AI agents.
The leaders’ calls for international safeguards contrasted with President Donald Trump’s opposition to broad global controls on AI.
- Who
- OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, Hugging Face co-founder Clément Delangue, Yoshua Bengio, and other officials and leaders.
- What
- They warned the UN Security Council about AI safety, control, concentration of power, and international security risks.
- When
- Wednesday, September 23, during the 2026 United Nations General Assembly; Trump gave his contrasting remarks the previous day.
- Where
- At a UN Security Council meeting held during the United Nations General Assembly.
- Why
- AI systems are advancing faster than many governments can regulate them, raising concerns about misuse, loss of control, and concentrated power.
International safeguards
Encouraging AI growth
Global rules
International safeguards
Altman, Amodei, Yoshua Bengio, and UN Secretary-General Antonio Guterres called for international cooperation or governance because AI risks cross borders and cannot be managed by one company or country.
Encouraging AI growth
Trump rejected what he called a globalist scheme to control AI and argued that broad international controls could stifle growth.
Speed of development
International safeguards
Amodei said companies should slow down as much as necessary to ensure successive AI systems are safe, while Altman urged extreme care.
Encouraging AI growth
Trump said the United States should encourage AI development rather than rein it in, while continuing to watch it closely through the Department of Justice.
Who should decide
International safeguards
Altman said AI's most important decisions should not be made solely by laboratories in San Francisco and should instead involve democratic processes and accountable governments.
Encouraging AI growth
The opposing approach described in the reports favors leaving more room for national leadership and market-driven development rather than imposing broad global controls.
- Meeting
- UN Security Council briefing convened by France during the annual General Assembly
- Main speakers
- Sam Altman, Dario Amodei, and Clément Delangue
- Amodei's warning
- Poorly managed AI could pose a risk to humanity as a whole
- Altman's position
- AI's most important decisions should be shaped through democratic processes and accountable governments
- Cybersecurity example
- Hugging Face said it used a Chinese AI model to defend against an AI-related attack
- Trump's position
- The United States should encourage AI growth rather than broadly rein it in
- International divide
- The United States and China have taken different approaches to AI oversight and global guardrails
