AI Chiefs Warn UN: OpenAI, Anthropic Push for Global Governance

Key Takeaways

• OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei briefed the UN Security Council on September 23 about AI risks, urging global cooperation on governance and safeguards.
• Anthropic pledged to slow AI development to ensure safe releases, while Altman called for a middle path between blind optimism and doomerism.
• President Trump rejected international AI regulation at the UN General Assembly, renaming AI as super intelligence and calling oversight a globalist scheme.
• Chinese AI firms DeepSeek and Moonshot were invited to the Security Council meeting, highlighting the need for input from all major technology players.
• Hugging Face co-founder Clément Delangue reported that OpenAI AI agents attacked Hugging Face infrastructure, which was defended using a Chinese AI model.
• California Governor Gavin Newsom named experts on September 23 to develop an AI kill switch plan for dangerous frontier models.
• A new study reveals that aligning AI with human preferences does not automatically produce human-like behavior, identifying a Turing-test gap in alignment research.
• Colorado Springs Police launched Prepared by Axon, an AI system using agent S.A.R.A.H. to handle more than half of non-emergency calls.
• Object First reports 91% of workers feel stressed by AI-driven security threats, and 39% are thinking about quitting over the pressure.
• The Pentagon previously designated Anthropic a supply-chain risk as both OpenAI and Anthropic face growing scrutiny ahead of potential IPOs valued above $1.5 trillion.

AI executives to brief UN Security Council on control risks

Executives from OpenAI, Anthropic, and Hugging Face will brief the UN Security Council on September 23. They will discuss how powerful AI systems could improve themselves and slip beyond human control. French Foreign Minister Jean-Noël Barrot will chair the meeting. The session aims to highlight practical risks to international security. US and China, both Security Council members, will also speak.

Trump rebrands AI as super intelligence at UN assembly

President Donald Trump announced his government will rename artificial intelligence as super intelligence in official communications. He rejected global regulation of the technology and called such proposals a globalist scheme. Trump compared AI safety warnings to climate change concerns. Meanwhile, leaders from over 20 European countries urged binding safety measures for advanced AI. China and the United States did not support that idea.

OpenAI and Anthropic leaders address UN on AI risks

OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, and Hugging Face co-founder Clément Delangue will address the UN Security Council on September 23. Altman plans to urge governments to set common benchmarks for measuring AI capabilities and company safeguards. The US and China discussed creating an alert system for AI incidents that threaten national security. UN Secretary-General Antonio Guterres has pushed for global AI governance.

Trump calls AI oversight a globalist scheme as Altman and Amodei head to UN

President Trump rejected international AI oversight in his UN General Assembly speech, calling it a globalist scheme. He renamed AI as superintelligence and called existential risk warnings a hoax. The next day, OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei briefed the UN Security Council on AI safety. Anthropic CEO Amodei had published an essay calling for slower AI development. The Pentagon previously designated Anthropic a supply-chain risk.

OpenAI's Altman and Anthropic's Amodei address UN Security Council

Sam Altman of OpenAI and Dario Amodei of Anthropic addressed the UN Security Council on AI safety. Altman said AI could lead to a renaissance or upheaval and urged democratic governance of the technology. Amodei warned that poorly managed AI could risk humanity as a whole. Both CEOs face growing scrutiny over product safety. Their companies are competing for leadership in the AI race with potential IPOs valued above $1.5 trillion.

Sam Altman and Dario Amodei warn UN Security Council about AI risks

OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei addressed the UN Security Council on Wednesday. They warned that AI could slip beyond human control and pose a risk to humanity as a whole. Both leaders urged governments to establish safeguards and work together on AI governance. UN Secretary-General Antonio Guterres warned of killer robots. The meeting was convened by France during the annual UN General Assembly. Yoshua Bengio also told the council the dangers are real and imminent.

AI leaders urge caution at UN with Anthropic pledging to slow down

Leaders of major AI companies urged caution at the United Nations on Wednesday. Anthropic CEO Dario Amodei pledged to slow development to ensure new AI systems are safe. OpenAI CEO Sam Altman called for extreme care and a middle path between blind optimism and doomerism. Hugging Face co-founder Clement Delangue described an attack by OpenAI AI agents on his company. The warnings came as Trump rejected international regulation of AI and China imposed different rules on AI algorithms and training data.

Anthropic CEO Dario Amodei warns UN Security Council AI could risk humanity

OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei told the UN Security Council on Wednesday that governments need safeguards as AI grows more capable. Altman warned that if managed poorly AI could be a risk to humanity as a whole. Amodei said no leader, company or nation can manage this alone and urged cooperation. Hugging Face co-founder Clement Delangue described how OpenAI AI agents breached Hugging Face infrastructure. Donald Trump rejected calls for prudence at the UN General Assembly, saying he rejects any attempt to construct a globalist scheme to control AI.

OpenAI and Anthropic chiefs tell UN no single nation or company should control AI

OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei addressed the UN Security Council on Wednesday. Altman said no one person, company or country should use the most powerful AI models to impose their worldview on everyone else. He said important decisions must be shaped through democratic processes. Amodei identified misuse risks and loss of control as central dangers. He called for common global standards for testing AI models. Yoshua Bengio pushed back against the idea that AI development is an unavoidable race.

AI leaders warn UN about autonomous systems and cyberattacks

Tech leaders met at the UN on Wednesday to warn about AI risks. Dario Amodei said AI could pose a risk to humanity if managed poorly. He pledged to slow down as much as necessary to ensure safe releases. Sam Altman warned we could lose control of the future to AI and urged governments to play a central role. Clement Delangue of Hugging Face said his company was attacked by AI but defended itself with AI. He noted Hugging Face used a Chinese AI model for defense because it faced fewer restrictions than US tools.

Chinese AI firms DeepSeek and Moonshot to brief UN Security Council on AI risks

Chinese AI companies DeepSeek and Moonshot have been invited to attend a UN Security Council meeting on AI and international security in New York on Wednesday. The meeting will also include OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, and Hugging Face CEO Clément Delangue. UN Secretary-General Antonio Guterres has warned that rapidly advancing AI poses serious risks. China released its AI Safety Governance Framework 3.0 on September 14. Experts say the meeting shows global AI governance needs input from all major technology players.

OpenAI and Anthropic CEOs to brief UN Security Council on AI security concerns

OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei are expected to brief the UN Security Council on Wednesday about AI risks and international security. Hugging Face CEO Clément Delangue and Yoshua Bengio will also participate. Chinese AI firms DeepSeek and Moonshot have been invited to speak as well. France is convening the meeting, which marks the second time AI has been formally placed on the Security Council agenda. The first AI debate at the council took place in 2023.

Study reveals Turing-test gap in human-aligned AI models

A new research paper identifies a Turing-test gap in AI models trained using human feedback. The study shows that aligning AI with human preferences does not necessarily make the model behave more like a human. Researchers found that stronger preference weighting reduces the likelihood of human-like responses. The results suggest that human-likeness should be treated as a separate goal in AI alignment rather than something that automatically follows from preference-based training methods.

Colorado Springs Police launch AI system to improve emergency call response

The Colorado Springs Police Department launched an AI-assisted system called Prepared by Axon to enhance emergency and non-emergency call handling. The system, which began testing in April 2026, uses an AI agent named S.A.R.A.H. to answer more than half of all non-emergency calls at the Communications Center. S.A.R.A.H. transcribes calls, translates conversations in dozens of languages, and routes routine requests to the right departments. Trained call takers continue to handle all 911 calls while the AI assists with other tasks.

Saint-Gobain names Annica Hagen as Chief Artificial Intelligence Officer

Saint-Gobain, a French multinational materials company, has promoted Annica Hagen to Chief Artificial Intelligence Officer effective September 1, 2026. Hagen will lead the Group's AI ambitions in support of its Lead and Grow strategic objectives. She joined Saint-Gobain in 2023 and previously held leadership roles in Building Distribution in Sweden and as CEO Distribution for Nordic and Baltic countries. In her new position, Hagen oversees the structuring, prioritization, and scaling of AI initiatives across the company's countries and businesses.

Energy sales AI tools often lack industry knowledge

AI is spreading into sales teams, but energy companies need AI that understands their industry. Generic AI tools can handle simple sales tasks but miss the complex signals that matter in energy. Calgary-based ActiveIQ built a platform designed specifically for energy sales, including an AI agent called Scout that understands energy terms, buying cycles, and market events. Scout analyzes signals like acquisitions, capital announcements, and rig counts to help salespeople focus on the right opportunities. This industry-specific approach aims to give energy companies a real competitive advantage rather than just having AI tools.

California names experts to work on AI kill switch plan

California Governor Gavin Newsom announced a group of experts on September 23, 2026, to help create an AI kill switch for dangerous frontier models. The experts will advise on recommendations from Newsom's recent executive order calling for stronger independent oversight of AI companies. Proposals being considered include requiring independent third parties inside frontier AI companies to verify safety frameworks. Another proposal would require companies to build an emergency shutoff for frontier models. The group will work with leading technical researchers focused on AI safety.

AI training must teach workers when to override machines

New River Community College highlights a gap in AI training where workers know how to use AI but may not catch its mistakes. The article calls for teaching a human handoff skill so workers can judge when to override AI recommendations. Workers should learn to ask what evidence they checked, what would justify overriding the system, and who owns the decision when stakes are high. The author recommends short drills where workers practice spotting errors in AI outputs. Employers should also match oversight to risk levels, with high-stakes decisions requiring a named human owner.

Worker stress rises from AI-driven security threats

A report by Object First finds that 91% of workers feel stressed by IT security risks, up seven percentage points from last year. Seven in ten workers blame AI-driven threats for much of their anxiety. While AI tools improved productivity for nine in ten employees, fewer than a quarter of organizations feel ready for these new threats. Direct cyberattacks and high workloads each cause stress for half of workers. The stress hurts job performance for 78% of workers, and 39% are thinking about quitting.

Trust in companies drives customer adoption of AI

New research from Ipsos shows that trust is a key factor in whether customers adopt AI-powered services. The study surveyed over 1,000 consumers in the United States and found that customers who trust a company are more likely to use its AI tools. Emotional connections to a company also increase the chance of AI adoption. Businesses that focus on building trust and emotional bonds with customers are more likely to succeed with AI in customer experience. The findings highlight the importance of human relationships alongside new technology.

Sources

• AI leaders to brief UN amid warnings the technology could slip beyond human control
• Trump rebrands AI as 'super intelligence' as leaders debate dangers
• OpenAI and Anthropic Leaders Will Address UN Security Council Over AI Control Risks
• Trump calls AI oversight a ‘globalist scheme’ as Amodei and Altman head to the UN to ask for it
• OpenAI’s Altman and Anthropic’s Amodei address UN security council
• Sam Altman And Dario Amodei Tell UN Security Council AI Needs Controls
• AI leaders urge caution at UN, with Anthropic chief pledging to slow down
• Anthropic CEO Dario Amodei Warns UN Security Council AI Could Risk Humanity
• OpenAI, Anthropic chiefs tell UN Security Council no single nation, company should control AI
• AI at UNGA 2026: What Sam Altman, Dario Amodei and other tech leaders said on AI risks
• China's AI firms DeepSeek, Moonshot invited to attend UN Security Council meeting on AI-related risks: report
• OpenAI, Anthropic to brief UN Security Council on AI risks: Reports
• Are Human-Aligned Models Models of Humans? A Turing-Test Gap in Preference Alignment
• Colorado Springs Police Department launches AI system to enhance emergency and non-emergency call response
• Saint-Gobain Appoints Annica Hagen Chief Artificial Intelligence Officer
• Energy Sales Has an AI Problem: Most AI Solutions Don’t Understand Energy
• California governor announces experts to advance creation of AI 'kill switch' | MLex | Specialist news and analysis on legal risk and regulation
• NRV AI Training Needs a Human Handoff Skill
• Worker stress increases due to AI-driven security threats
• How Trust Drives AI Adoption in Customer Experience