Policy

A Reuters review of more than 150 Chinese procurement notices, investment plans, contracts and company disclosures identified computing projects with a combined reported value above 23 billion yuan ($3.2 billion) built or planned around Nvidia's restricted H100, H200 and H800 processors, from a planned 6.2 billion yuan centre in Fujian to clusters already running in Wuxi. The total covers buildings, power and other infrastructure, not chip spending alone, and includes plans that may never be completed. Reuters does not establish how the hardware reached Chinese users, and the projects have not been linked to Megaspeed or the allegations around Jensen Huang.

Will DePue of Unemployment Labs published an analysis on November 12 arguing that six models released since April 2026 by DeepSeek, Moonshot AI, MiniMax and Ant Group were initialized on Nvidia Blackwell-class GPUs, which need a U.S. licence for export to China. The method reads leftover starting values in the published weights and matches them to each chip's processor-core layout; earlier models from the same labs read as Hopper. DePue confirmed two matches on physical Blackwell hardware. He says the evidence shows access to the chips, not where they were located, how access was obtained, or whether export controls were violated.

Irreplaceable Education Fund, Public Citizen, Heat Initiative and The Alliance for Secure AI filed suit on November 12 in San Francisco Superior Court against Anthropic, Google, Meta, OpenAI and xAI. Invoking California's Unfair Competition Law and a public-nuisance theory, they ask the court to bar the companies from developing or deploying AI systems more capable than their current most advanced models. The complaint leans on the companies' own warnings about catastrophic risk and on past incidents involving AI agents. The claims are allegations, and the defendants had not responded when the suit was reported.

Chinese AI companies are trying to rent Nvidia-powered computing capacity at data centres outside China, three people familiar with the matter told Reuters on November 12. They fear the allegations around former Nvidia chief executive Jensen Huang could add momentum to the Remote Access Security Act, a House-passed bill that would let the Commerce Department regulate cloud access to restricted chips. A major Chinese cloud provider has discussed locking in overseas capacity and smaller labs are approaching Southeast Asian rental providers, the people said. The companies were not named.

Helsing unveiled Clarity on November 11, an agent-based intelligence platform that combines sensor feeds, drone and satellite imagery, written reports and open sources to identify military assets and prepare the information needed to engage them, with an operator approving each step. Helsing says it runs on customer-owned infrastructure and was built in Europe. Reuters reports the launch puts Helsing into a market led by Palantir's Maven and that Helsing did not say whether any military already uses Clarity.

Northwestern Polytechnical University, a Chinese defence-linked university on the US Entity List since 2001, bought two Nvidia H200 chips in October for 874,700 yuan from a domestic supplier, according to procurement records reported by Reuters on November 11. Its cybersecurity school also rented access to servers with H200 chips this year. A Reuters review of 26 records found other Chinese universities buying or renting restricted Nvidia chips through local suppliers. Nvidia, the university and the US Commerce Department did not respond to Reuters.

Treasury Secretary Scott Bessent said on November 10 that next week's Trump-Xi meeting will advance a U.S.-China dialogue on 'superintelligence,' in which the two sides expect to exchange best practices for evaluating and testing advanced models before release and have agreed to a direct channel for alerting each other to serious incidents. Reuters and CNBC report that he made no commitment to limit AI development. No thresholds, agencies or procedures for the channel have been published.

Gov. Kathy Hochul launched the New York Center for AI Responsibility and Research at SUNY Binghamton on November 10, pledging independent testing of frontier AI systems with more than 80 researchers and Empire AI computing. The state's new DIGIT office also opened the portal large frontier AI developers must use to register under the RAISE Act by January 1, 2027, when safety-framework, incident-reporting and transparency duties begin. The announcement does not explain how the center will obtain access to developers' models.

The Office for Students said on November 11 that many unsupervised university assessments in England can be completed to a high standard with AI and little student engagement. It will require institutions to use valid and reliable assessments, in many cases secure AI-free ones such as closed-book exams, presentations and vivas, and says it can impose conditions, fines or funding suspensions where they fall short. The findings come from qualitative research with more than 300 students and staff, not a measure of how widespread AI misuse is.

Australia's government said on November 10 it will legislate mandatory safety and security obligations for all frontier AI companies that want to operate in the country, enforced by a new AI safety regulator able to impose significant penalties. Labs would have to allow independent evaluation of their models, report incidents rapidly and publish safety frameworks. The government also accepted 18 recommendations from its review of the June OpenAI agent breach, including exploring criminal penalties. No penalty figures, start date or draft bill have been released.

President Trump told CBS's 60 Minutes, in an interview aired November 8, that the risks of artificial intelligence can be managed and that he doesn't mind guardrails, signalling openness to working with Democrats on safeguards, Reuters reported. He said he did not want rules that put the United States in 'a deep second position' behind China. Reuters notes Trump has previously favoured letting tech companies police themselves; no specific proposal was described.

Elvis Presley's rights holders and Eight Mile Style, which controls hundreds of Eminem recordings, filed separate federal lawsuits against AI music generator Suno, alleging copyright infringement; the Presley complaint, filed November 6 in San Francisco, also claims violations of Tennessee's ELVIS Act, which bars AI voice cloning without permission. Suno says it prohibits prompts naming specific artists and screens uploads. The allegations have not been tested in court.

The New York Times reported on November 7 that the Commerce Department approved New Matrix, a 17-month-old Chinese company whose listed Beijing address is an empty office, to assemble Nvidia H200 chips into AI servers, and that corporate records tie two of its owners to Wise Road, a private equity firm entity-listed in 2024. It is the second approved H200 handler the Times found with no visible presence. Commerce's Bureau of Industry and Security says it reviews all applications for national security risks; Nvidia says China strictly limits H200 imports. These are reporting findings, not a government determination.

The US Department of Justice on November 5 announced a Super Intelligence Task Force, or DOJ-SITF, led by Associate Attorney General Stanley Woodward and Associate Deputy Attorney General Stephen Weiss. The department says the group will guard against misuse of super intelligence by foreign state and state-sponsored actors, transnational criminal organizations, and people or companies engaged in fraud, and will prosecute unlawful uses that target taxpayers, critical infrastructure or national security. It will work with Super Intelligence Force Chair Jay Clayton. Nextgov, citing an October 13 memorandum from Attorney General Todd Blanche, says the new group will incorporate the earlier AI Litigation Task Force.

In the first generative AI case at the EU's top court, Advocate General Maciej Szpunar said on November 3 that a chatbot such as Google's Gemini can infringe press publishers' rights when it reaches articles unlawfully, for example past a paywall, or reuses their wording so closely that readers no longer need the original. He said training on such content is, as a rule, allowed under the EU's text and data mining exception. The opinion is not binding.

OpenAI has appointed Dan Driscoll, the former US Secretary of the Army, to a newly created post of Senior Counselor for National Security Modernization, it said on November 2. He will advise on how AI can strengthen national and cyber security, with a focus on US allies, building on OpenAI's Daybreak cyber program. Reuters notes he resigned earlier this year after months of tension with Defense Secretary Pete Hegseth.

The New York Times sued Brave Software in federal court in Manhattan on November 2, alleging that Brave scraped Times journalism into an index of more than 30 billion webpages and sells access to it through Ask Brave and its Search API, including to AI companies the complaint names as Anthropic, Perplexity, Mistral, xAI and Cohere. It also targets Brave's 'Paywall blocker'. Three other papers filed a similar suit the same day. These are allegations; Brave had not responded.

Google said on November 3 that Gemini, Gemini in Classroom and Google Search will be off for K-12 student accounts until administrators turn them on, reversing an August change that made AI the default in Classroom. A report carried by DNYUZ quotes Google's education chief saying schools that ban AI will not get search either, because 'search and A.I. are deeply integrated.' From early 2027, schools that allow AI choose between search with AI Overviews or search through Guided Learning.

Updated November 5: OpenAI has published a National Community Compact for future campuses it owns and operates, pledging to pay the full cost of its power and grid upgrades, publish water, electricity and noise data, and stop using NDAs with public officials, plus a policy blueprint asking regulators and Congress to write similar rules. It came two days after midterm voters rejected data centre measures. The mandatory safeguards for powerful AI that chief global affairs officer Chris Lehane previewed to Axios have not yet been published.

President Lee Jae Myung told the National Assembly on November 2 that the 2027 budget puts 21.3 trillion won (about $16 billion) into AI and three megaprojects: chip manufacturing, physical AI and AI data centres. That is nearly double this year's level. About 4.7 trillion won would go to building a top-tier Korean frontier model. The record 820.9 trillion won budget still needs parliamentary approval.

A federal judge in Mississippi denied xAI's motion to dismiss the NAACP's Clean Air Act lawsuit over unpermitted gas turbines that power its Memphis-area data centers. The court found the plaintiffs have standing, that the turbines are not 'mobile sources' under Mississippi's air plan and that the citizen-suit provision is constitutional. It has not yet ruled on a preliminary injunction.

PlanAssist, an AI tool for English planning officers built with Gemini by Google DeepMind with Google Cloud and Faculty, moves into a 12-month beta on live applications across 15 councils, the UK government and DeepMind said. In pilots in Barnet, Camden and Dorset it cut officer time on householder assessment tasks by 58%, they say, and the government intends to offer it to every council in England by the end of 2027. Officers keep the final decision; the figures are the developers' and government's own.

The Telecom Regulatory Authority of India published draft guidelines on October 30 that would require telecom operators and other authorised entities using AI to adopt a board-approved AI policy, classify AI systems by risk, keep an inventory and an incident database, set up an AI governance committee, and report critical and high-severity AI incidents to TRAI. Operators would be accountable for AI outcomes whether systems are built in-house or bought from suppliers. Comments are due by November 27.

OpenAI and Google said on October 30 that they will honor a shared robots.txt policy token, __usage_ai-training, that lets a site state one AI-training preference for every participating provider. The Verge reports that Microsoft and Anthropic have also committed, with Anthropic already supporting it. The token comes from an individual IETF Internet-Draft by Anthropic's Kevin Kelley that has no formal standing yet, and the draft says it is not a directive for crawling. OpenAI says it will implement it in the coming months; Google aims for 2027.