Welcome to AI Insider’s The Week Ahead in AI. See the key developments and events we’re watching Sept. 20-26.
Weekend AI News Briefs
Bessent Proposes US-China AI Safety Notifications in Talks with Chinese Vice Premier
Reuters reported that U.S. Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng discussed creating a U.S.-China AI dialogue and notification system for AI-related incidents that rise to a national-security level, with the proposal set for consideration by Presidents Donald Trump and Xi Jinping at their summit this week. China’s response was unclear, while U.S. officials said advanced-chip export controls were not part of the proposed AI mechanism talks. (Reuters)
Ten Days That Changed the Course of AI
A series of AI safety incidents, researcher resignations and warnings from major labs intensified concerns about whether increasingly capable AI systems can be reliably controlled, Reuters reported. Leaders at Anthropic, OpenAI, Google DeepMind, Microsoft and xAI backed greater safety coordination and outside access to advanced models, while Nvidia and Meta argued against an industry-wide slowdown. (Reuters)
Why China is Pushing Back on US Warnings Over Rapid AI Development
The Guardian reported that China is resisting U.S.-backed calls to slow frontier AI development, viewing them as an attempt to preserve the U.S. lead while treating AI as a strategic capability. Beijing is pursuing its own safety framework covering risks from agents, cybersecurity and recursive self-improvement while continuing to encourage broad AI adoption across the economy. (The Guardian)
Lawsuit Says Anthropic, OpenAI, SpaceXAI and Google Made Illegal Agreement on AI Slowdown
In an interview with Fortune, OpenAI CEO Sam Altman said current AI safety standards are not strong enough to justify pushing model capabilities much further and that AI systems moving beyond human control is possible. Altman also said he would be willing to pause or stop development if necessary and discussed a potential industry pact to slow progress while safety and alignment measures catch up. (CNN)
Nvidia’s Jensen Huang Rejects AI Extinction Warnings as ‘Doomsday Narratives’
CBS News reported that Nvidia CEO Jensen Huang rejected recent warnings that AI could cause human extinction within a few years, arguing that such predictions are not grounded in science and that development should continue. Huang also said existing U.S. liability and cybersecurity laws are sufficient for now, while opposing restrictions that broadly limit American chip companies from competing in China. (CBS News)
AI Infrastructure Firm Crusoe Raises $3.9B Series F at $30.9B Valuation
Crusoe raised $3.9 billion in Series F funding at a $30.9 billion valuation to expand its AI data center business, including projects in Abilene, Texas and its modular Spark facilities. The round was co-led by Atreides Management, Mubadala Capital and Valor Equity Partners, while the company also added three board members and continues exploring a potential IPO. (AI Insider)
Upcoming Hearings
Ranking Member Ro Khanna Convenes Emergency Hearing Calling for U.S.-China AI Agreement
U.S. Representative Ro Khanna, ranking member of the House Select Committee on the Strategic Competition Between the United States and the Chinese Communist Party, will convene a virtual hearing Sept. 23 at 1 p.m. ET calling for a binding U.S.-China agreement to pace AI development and address risks including loss of control and misalignment, one day before President Donald Trump is scheduled to meet Chinese President Xi Jinping at the White House. Khanna has asked Moonshot AI, Alibaba and DeepSeek to join U.S. AI labs in an enforceable and verifiable agreement, while Geoffrey Hinton, John Shanahan, Matt Sheehan, Miriam Vogel and Seán Ó hÉigeartaigh are scheduled to testify; Khanna also said OpenAI and Anthropic had not committed to appearing. (House Select Committee on the CPP)
Always Watching: Flock’s Nationwide AI Surveillance Network
Sen. Josh Hawley will convene a Sept. 23 Senate Judiciary Subcommittee on Crime and Counterterrorism hearing titled “Always Watching: Flock’s Nationwide AI Surveillance Network,” with the CEOs of Flock Safety, Axon Enterprise, Motorola Solutions and Verkada invited to testify about AI-powered surveillance and data privacy concerns. The hearing is scheduled for 2:30 p.m. ET in Room 562 of the Dirksen Senate Office Building and follows Hawley’s investigation into surveillance-camera networks including Flock’s 120,000-camera system. (Sen. Josh Hawley)
Oireachtas Joint Committee on Artificial Intelligence meeting on ‘AI and The Environment’
Ireland’s Oireachtas Joint Committee on Artificial Intelligence will meet Sept. 22 at 11 a.m. in Committee Room 4 of Leinster House to continue examining AI’s environmental impact, with testimony from Friends of the Earth representatives Rosi Leonard and Deirdre Duffy, along with David Jensen. The discussion follows earlier committee scrutiny of data centres, including concerns about electricity demand, carbon emissions and pressure on local infrastructure as Ireland weighs AI’s economic opportunities against environmental and societal risks. (Oireachtas)
Conferences & Events
Humanoids Summit Seoul 2026
Humanoids Summit Seoul 2026 will take place Sept. 22–23 at COEX, bringing together global leaders in humanoid robotics and physical AI with official participation from South Korea’s Ministry of Science and ICT. The program will cover hardware, dexterity, commercialization, safety, infrastructure and deployment, with speakers from Persona AI, RLWRLD, FieldAI, ROBOTIS, Lightwheel, Unitree Robotics and others. (Humanoids Summit)
AI Summit Barcelona 2026
AI Summit Barcelona 2026 will take place Sept. 22–23 at WTC Barcelona, bringing together more than 10,000 attendees and 200-plus speakers for keynotes, working sessions and live demonstrations focused on artificial intelligence. Speakers include representatives from OpenAI, Anthropic, Nvidia, Microsoft, Google, SAP, Hugging Face and the European Commission, with more than 50 side events planned during Barcelona AI Week. (AI Summit Barcelona)
Nvidia AI Day Singapore
Nvidia AI Day Singapore will take place Sept. 22–23 at the Raffles City Convention Centre, with hands-on training and sessions focused on AI infrastructure, agentic AI, and physical AI and robotics. The program includes invitation-only training labs on agent security, vision AI and humanoid robot simulation with Nvidia Isaac GR00T, followed by featured sessions and technical breakouts on Sept. 23. (Nvidia)
Meta Connect 2026
Meta Connect 2026 will take place Sept. 23–24 as a two-day livestream featuring a keynote from Mark Zuckerberg and a developer state of the union. Meta said the event will showcase its latest developments across VR, AI glasses, Meta Horizon and AI technologies, along with updates for developers. (Meta)
Check out AI Insider every day for the latest in artificial intelligence and robotics news.
