US and Chinese officials wrapped up talks in New York on September 20, with Treasury Secretary Scott Bessent calling the discussions very successful. The two sides agreed to set up an official AI dialogue and a security mechanism for sharing early warnings about national security-related AI incidents. Bessent proposed the system for Presidents Trump and Xi to consider, though export controls on AI chips were not discussed. A working group will continue the conversations on Monday, and future talks will address weaponization of AI and cyberattack prevention.
President Trump is set to host President Xi Jinping at the White House later this week for the first state visit in over a decade. Trump rejects any deal that would slow AI research, while Senator Mike Rounds warned that China's rapid AI development poses a critical national security threat. Rounds urged a balanced approach to regulation that fosters domestic innovation rather than stifling it.
Trade topics also featured prominently in the US-China discussions. Both sides formalized a Board of Trade first discussed in Beijing in May, and Bessent said the meeting covered possible lower tariffs on non-sensitive goods. A truce deal on trade tensions nears its November expiration, adding urgency to the negotiations. France and Canada also sought to deepen ties with China on critical minerals and energy during the broader diplomatic push.
In the private sector, Belgian cybersecurity firm Aikido released an open-weight AI model called Altar on September 21, built on Z.AI's GLM-5.3 and tailored for cybersecurity use. Altar allows companies to run security tools locally without sending sensitive code to external providers. Aikido also launched Aikido-OS, designed to detect and prevent cyber attacks in real time. The company reached a $1 billion valuation in January and is deploying its tools with customers including Belgian bank Belfius.
New York State is advancing AI protections as the Trump administration pushes ahead with AI development. Governor Kathy Hochul and other officials are promoting measures including the RAISE Act and considering a kill switch for AI systems. Meanwhile, ArtPrize in Grand Rapids, Michigan, clarified that its 2026 competition does not allow entries created entirely with AI, though limited use of generative tools is permitted in the New Media category. Pope Leo XIV, speaking at the AI4Peace symposium in Rome, called for AI to serve human dignity and peace.
The US Coast Guard will open a Center for Artificial Intelligence and Machine Learning this fall at the Academy in New London, Connecticut. The center partners with Johns Hopkins, Brown, MIT Lincoln Laboratory, and the NSA, focusing on building practical AI tools for the fleet. Projects already underway include optimizing buoy positioning and improving cutter patrol schedules, with working prototypes targeted within six months. Cisco's 2026 Global AI Partner Study found that only 29% of partners report high awareness of AI-specific security threats, down from 35% last year, while nearly half of customers are piloting or scaling agentic AI.
Key Takeaways
- US and China agreed to an official AI dialogue with a security mechanism for sharing early warnings on national security-related AI incidents
- President Trump rejects any deal to slow AI research ahead of his summit with Xi Jinping at the White House
- US-China talks also formalized a Board of Trade and discussed lower tariffs on non-sensitive goods ahead of a November truce deal expiration
- Belgian firm Aikido released Altar, an open-weight AI model based on Z.AI's GLM-5.3, for local cybersecurity use, and reached a $1 billion valuation in January
- Cybersecurity buyers face paralysis from too many similar options, with 737 VC deals worth $12.94 billion globally in early 2026
- Senator Mike Rounds warns China's rapid AI development is a critical national security threat and urges balanced regulation
- New York State is advancing AI protections including the RAISE Act and considering a kill switch for AI systems
- The US Coast Guard will open an AI and Machine Learning Center in New London, Connecticut, partnering with Johns Hopkins, MIT Lincoln Laboratory, and the NSA
- Cisco's 2026 study shows only 29% of partners have high awareness of AI-specific security threats, while 92% report customer concerns about AI running costs
- ArtPrize in Michigan bars fully AI-created entries from its 2026 competition, while Pope Leo XIV calls for AI to serve human dignity and peace
US proposes AI safety alert exchange with China
The United States proposed a new system to share AI safety alerts with China. Treasury Secretary Scott Bessent announced this after talks in New York. The plan aims to increase transparency between the world's two largest AI powers. Both sides discussed setting up a U.S.-China AI dialogue for national security incidents. Experts doubt a major AI agreement will come from the upcoming Trump-Xi summit.
US Treasury calls China trade and AI meeting very successful
US Treasury Secretary Scott Bessent and Chinese Vice-Premier He Lifeng ended talks in New York on September 20. The US proposed a new AI safety notification system for Presidents Trump and Xi to consider. Bessent stressed the need for more transparency between the top two AI powers. The meeting also covered trade topics like a new Board of Trade and possible lower tariffs on non-sensitive goods. Future talks will tackle sensitive issues like weaponization of AI and cyberattack prevention.
US and China discuss AI dialogue and trade ahead of summit
US and Chinese delegations held constructive talks in New York on Sunday. The discussions covered artificial intelligence and trade before President Xi Jinping's visit to Washington. Both sides discussed setting up an AI dialogue mechanism due to rising security fears. They also prepared to ease trade tensions as a truce deal nears its November expiration. France and Canada also sought to deepen ties on critical minerals and energy.
Bessent praises China talks on AI and trade
US Treasury Secretary Scott Bessent called the talks with China very successful. The talks spanned AI, trade, and investment before the Trump-Xi summit. Both sides agreed to set up an AI dialogue to understand goals and threats. A working group will continue discussions on Monday. President Trump is set to host President Xi at the White House later this week for the first state visit in over a decade.
US and China agree on AI dialogue with security mechanism
The US and China agreed to an official AI dialogue to build more transparency. Treasury Secretary Scott Bessent proposed an early-warning system for security-related AI incidents. Export controls on AI chips were not part of the talks. President Trump rejects any deal to slow AI research. The two sides also formalized a Board of Trade first discussed in Beijing in May. Trump will welcome Xi at Joint Base Andrews on Wednesday for a summit in Washington.
Belgian firm Aikido releases open-weight AI model for cybersecurity
Belgian cybersecurity firm Aikido released an open-weight AI model called Altar on September 21. Altar is a customized version of Z.AI's GLM-5.3 model designed for cybersecurity use. It allows companies to use AI security tools locally without sending sensitive code to external providers. The model will be deployed in solutions for customers including Belgian bank Belfius. Aikido reached a $1 billion valuation in January.
Aikido debuts Aikido-OS AI model to detect cyber attacks in real time
Belgian cybersecurity firm Aikido launched an AI model called Aikido-OS designed to detect and prevent cyber attacks in real time. The model uses machine learning algorithms to analyze data from various sources and identify potential threats. Demand for local cybersecurity tools has grown due to concerns over data privacy and security. Aikido-OS is the first AI model of its kind developed in Belgium. The company aims to lead the European cybersecurity market.
GOP lawmaker Rounds warns US must lead in AI development
Senator Mike Rounds warned that China's rapid artificial intelligence development poses a critical national security threat. He spoke during an interview on Life, Liberty and Levin. Rounds said the United States must outpace foreign adversaries in AI technology to secure future global leadership. He urged a balanced approach to regulation that fosters domestic innovation.
AI fears boost security budgets but buyers struggle to choose
Growing fears about AI attack capabilities are opening up enterprise security budgets as executives wake up to the threats. However, buyers are paralyzed by the prospect of choosing the wrong solution from many similar options. In 2025, SYN Ventures met with 559 cybersecurity companies it had never met before. Globally, cybersecurity startups closed 737 VC deals worth $12.94 billion in the first part of 2026. Many companies are features rather than complete products, making it harder for buyers to decide.
US Coast Guard opens AI and machine learning center in Connecticut
The US Coast Guard will open a new Center for Artificial Intelligence and Machine Learning this fall at the US Coast Guard Academy in New London, Connecticut. The center links the Coast Guard to research institutions including Johns Hopkins University, Brown University, MIT Lincoln Laboratory, and the National Security Agency. It focuses on three goals: building fast practical AI tools for the fleet, conducting mission-focused research, and training the workforce for AI use. Projects already underway include optimizing buoy positioning and improving cutter patrol schedules. The center aims to build working prototypes within six months.
Cisco study shows AI market maturing as security concerns grow
Cisco's 2026 Global AI Partner Study found that only 29% of partners report high awareness of AI-specific security threats, down from 35% last year. Nearly half of customers are now piloting or scaling agentic AI, but multi-agent orchestration remains the weakest capability. About 92% of partners say customers have raised concerns about AI running costs, while only 43% say customer ROI is meeting expectations. Consulting and advisory services have jumped from 12% to 50% as a leading growth opportunity for partners.
ArtPrize sets rules for AI art in 2026 competition
ArtPrize organizers in Grand Rapids, Michigan said they do not allow any entries created entirely with AI. The New Media category allows limited use of generative AI tools but only recognizes humans as artists and designers. Some attendees supported labeled AI categories while others called AI art fake or cheap. ArtPrize also clarified that its voting partner FlockX is not connected to Flock Safety and is only used to cast votes.
Pope Leo XIV says AI must serve human dignity and peace
Pope Leo XIV called for AI development to be guided by wisdom and charity so it can serve human dignity and strengthen human bonds. His message was read at the AI4Peace symposium at the Pontifical Lateran University in Rome. He encouraged participants to explore how AI can foster dialogue, reconciliation and solidarity among peoples. The Pope stressed that technological progress should be placed at the service of the common good.
New York law boosts AI protections as Hochul eyes kill switch
New York State leaders are focusing on strengthening AI protections as the Trump administration announced plans to continue AI development. Governor Kathy Hochul and other officials are advancing measures including the RAISE Act. State officials are also considering a kill switch for AI systems. The push comes amid growing concerns about unregulated AI development.
Engineer's viral post highlights dark side of AI coding
A software engineer's viral post says AI coding tools are turning engineers into button-pushers. The post highlights the soul-crushing downside of relying on AI for coding tasks. Morteza Nikoubazl photographed the story for NurPhoto via Getty Images. The viral message has sparked discussion about how AI is changing the engineering profession.
Sources
- U.S. proposes exchanging AI safety alerts with China, Bessent says
- US Treasury chief says meeting with China on trade, AI ‘very successful’
- Business - US, China discuss AI dialogue mechanism and trade ahead of Trump-XI summit
- Bessent hails ‘very successful’ China talks on AI, trade
- US and China agree on AI dialogue with security mechanism ahead of Trump-Xi summit
- Belgium's Aikido launches cybersecurity AI model as demand for local tools grows
- Belgium's Aikido launches cybersecurity AI model as demand for local tools grows
- GOP lawmaker warns US must outpace China in artificial intelligence
- AI fear is filling security budgets. Buyers still can't decide.
- US Coast Guard launches AI and machine learning center to advance operational readiness, workforce development
- Partner Study Shows a Maturing AI Market, Expanding Opportunity for Partners
- ArtPrize addresses AI use, voting system for 2026 competition
- Pope Leo XIV: AI must serve human dignity and strengthen peace
- This new NY law increases AI protections as Hochul eyes 'kill switch'
- Engineer's viral post captures the soul-crushing downside of AI coding
Comments
Please log in to post a comment.