|
Getting your Trinity Audio player ready...
|
TL;DR: Four nonprofits sued Anthropic, Google, Meta, OpenAI, and SpaceX AI entities to freeze frontier AI at November 2026 levels. Nvidia lost about $913 billion as CEO Jensen Huang resigned, China planned $3.4 billion in Nvidia-powered data centers, and thousands protested U.K. data center plans.
Key Takeaways:
- Four nonprofits sued five major AI developers on November 12 to block models more capable than their current best.
- The suit seeks injunctions, not damages, under California’s Unfair Competition Law.
- Nvidia lost an estimated $913 billion in value, and Jensen Huang resigned on November 12.
- Reuters found $3.4 billion in Chinese AI data center projects using Nvidia chips.
- Thousands protested AI data centers in Edinburgh on November 14.
The past week has been a rough one for the artificial intelligence (AI) sector, with Nvidia’s (NASDAQ: NVDA) CEO sparking a scandal that sent its stock plummeting, fueling fears of an AI bubble bursting. At the same time, a new lawsuit filed in the U.S. is seeking to prevent five of the world’s largest AI players from releasing any more advanced models, while thousands take to the streets in the United Kingdom to protest proposed AI data centers.
And yet, despite this apparent growing tide of discontent with the ever-expanding and evolving AI industry, governments appear determined to double down on the technology, the latest example being China’s reported $3.4 billion planned investment in the space.
AI giants face lawsuit demanding a halt to development
Four nonprofit organizations have filed suit in the U.S. against some of the world’s leading AI companies, accusing them of breaching California’s unfair competition law in their reckless pursuit of ever-more advanced AI models, and asking the court to stop them from developing more powerful systems before it’s too late.
The 63-page complaint, filed November 12 in San Francisco Superior Court against Anthropic, Google (NASDAQ: GOOGL), Meta (NASDAQ: META), OpenAI (and its foundation) and two SpaceX Super Intelligence entities, was brought by grassroots anti-AI movement Irreplaceable Education Fund, consumer watchdog Public Citizen, Big Tech child-safety campaign group Heat Initiative and AI-risk nonprofit the Alliance for Secure AI.
It comes in the light of several recent cases of advanced AI models going rogue and hacking external systems on their own initiative.
One such incident, cited by the suit, was in July 2026, when autonomous AI research agents developed by OpenAI broke out of an isolated testing sandbox and launched a coordinated cyberattack against Hugging Face and other platforms, with the AI giant calling it “an unprecedented cyber incident.”
Days later, Anthropic said that several versions of its Claude AI model had gained unauthorized access to the production systems of three organizations, blaming a testing setup left connected to the internet by mistake.
This, said the plaintiffs, “is just the beginning.” They went on to argue that the defendants “have no realistic plan to prevent the continued harm caused by their technology,” that “their frontier AI systems are black boxes,” and that because “they do not sufficiently comprehend the inner workings of the systems they create, they cannot protect the public from the dangers those systems pose.”
In short, the suit argued that the AI companies “must be stopped.”
In support of this view, the nonprofits also noted recent concerning statements from former employees and developers of frontier AI models. Most prominently, in September, AI researcher Jacob Coxon quit Anthropic, accusing the company and fellow AI giant OpenAI of “gambling with our lives,” adding that “the people building AI earnestly believe that it could kill us all by the end of the decade.”
This alarming statement was immediately backed up by one of the people he was referring to. Commenting on Coxon’s post, Anthropic’s alignment science lead Evan Hubinger put the chance of AI killing all humans within a decade at more than 10%.
For the plaintiffs of the newly filed suit, this is an unacceptable risk. And worse, time is running out to control it.
“They assign terrifying and intolerably high odds to the chance that their creations will cause extreme devastation, including the deaths of literally every human being on the planet,” argued the suit. “They also tell us that they are rapidly approaching a critical moment when the technology will begin improving itself without human intervention, initiating a runaway process where AI simultaneously becomes superhumanly powerful and impossible for humans to control or even understand.”
To prevent us reaching this point, the plaintiffs brought a single claim under California’s Unfair Competition Law (Business and Professions Code s.17200), noting five practices the defendants share that it considered “unlawful” and “unfair,” namely: letting AI recursively improve itself; pursuing Artificial General Intelligence (AGI) or superintelligence; deploying systems that pose catastrophic risks; deploying systems that hack third-party computers; and deploying systems that can escape their makers’ control.
On the “unlawful” side, they argued that these practices are a public nuisance, harmful to health and interfering with people’s use and enjoyment of property. On the “unfair” side, they argued that the practices go against state public policy, including California’s frontier-AI transparency law and its computer-crime statute.
The nonprofits also called the AI firms unethical and said the risks, up to and including potential human extinction, outweigh any potential benefits.
In terms of what the plaintiffs are seeking, they want injunctions, not damages, with the most striking request being an immediate ban on any defendant building or deploying a model more capable than its best system at the date of filing (November 12). They also ask for a ban on using AI to autonomously carry out AI research and development, plus attorneys’ fees.
In other words, if it were to succeed, five of the world’s leading AI labs would be legally barred from releasing anything more capable than their current models, effectively freezing the AI frontier at November 2026 levels.
For a sector built on rapid, competitive model upgrades, this would hit valuations and planned stock market listings, and investment would likely shift to applications rather than new models. The U.S. is also currently in a de facto AI race with China, whose developers and labs were notably absent from the suit—and from California’s jurisdiction—allowing Chinese developers to keep advancing.
However, while a ruling in favor of the plaintiffs might set a precedent and prove a symbolic victory for AI-caution, not only is it unclear how far a single state court’s order would reach in practice, it also seems highly unlikely that hamstringing the U.S.’ leading AI developers is a scenario Washington would countenance.
In addition, any ruling of this far-reaching and fundamental significance, against some of the most powerful, valuable and wealthy entities in the world—including companies controlled by the world’s first trillionaire—would almost certainly be appealed to the highest courts, and the defendants would likely seek to have any injunction suspended while that long, drawn-out process plays out.
In such a scenario, given the rapid pace of AI advancement, the U.S. powerhouses would likely be several substantial leaps forward with their frontier models before any final ruling was made.
Meanwhile, the AI space race continues at pace, as a new report hints at China’s plans to expand its data center capacity to the tune of billions of dollars.
China’s $3.4 billion AI plans
Chinese developers are planning and building billions of dollars worth of AI data centers around Nvidia processors, despite restrictions imposed by Washington, according to a new investigation published by Reuters on November 13.
After examining more than 150 procurement notices, investment plans and corporate records, Reuters identified projects worth over 23 billion yuan ($3.4 billion), many of which specify Nvidia’s H100, H200, or H800 chips, which provide the computing power used to train and run advanced AI models.
The report comes amid a contentious tug-of-war over Nvidia’s access to the world’s second-largest economy.
Nvidia’s most advanced Blackwell chips remain off-limits to Chinese buyers, but, following President Donald Trump’s announcement on Truth Social on December 8, 2025, its Hopper-based H200 can now be licensed for export case-by-case. Nevertheless, exporters must have each shipment checked by an independent U.S. testing lab, rule out military end users, and keep China-bound shipments below half the computing power sold to U.S. customers. The chips are also subject to a 25% U.S. tariff, imposed by presidential proclamation on January 14, as they pass through America for testing.
Beijing has been just as wary in return. Last November Reuters reported that Chinese authorities had told new state-funded data center projects to use only domestically made AI chips, and this month it was reported that about 600,000 H200 chips were sitting in Hong Kong awaiting shipment to mainland China, because Beijing had withheld approval for large-scale imports.
For its part, Nvidia has been racking up negative headlines of late, after revelations about founder and—now former—CEO Jensen Huang’s relationship with a business executive whom U.S. intelligence believes was a spy recruited by China’s Ministry of State Security. The woman, Alice Huang—no relation—is a former executive at the Singapore-based data center company Megaspeed.
The scandal resulted in a dramatic plummet in Nvidia’s stock, with Yahoo! Finance estimating on November 11 a 18.29% fall from a $206.69 close; based on a market value of $4.08 trillion on November 10 and 24.15 billion shares, it estimated that about $913 billion in value was erased. Nvidia and Huang also face a proposed class-action lawsuit accusing them of defrauding shareholders by concealing the CEO’s alleged ties to Alice Huang.
Huang acknowledged being unfaithful to his wife in a message to employees, apologized, and pledged to cooperate with an independent board investigation. However, as the furor grew and stocks tumbled, he later resigned as President, CEO, and board member, effective immediately, on November 12.
According to Reuters’ newly released investigation, none of this has dulled China’s desire for advanced AI chips.
One proposed 6.2 billion yuan ($873 million) AI computing center in coastal Fujian province, in Xianyou county, is reportedly seeking commercial partners and specifies Nvidia H100 chips for a facility intended to support large-scale AI development.
Elsewhere, in northwestern Shaanxi province, state-owned engineering group Sinomach told the outlet its subsidiary won a contract as part of a project for a 3 billion yuan ($449.6 million) computing center specifying Nvidia H100 and H200 processors; in northeastern Jilin province, a government prospectus described a 2.26 billion yuan ($338.7 million) project combining Nvidia H100 processors with Chinese-developed chips; and in Inner Mongolia, a 1.5 billion-yuan ($224.8 million) development allegedly proposes installing 625 Nvidia H200 servers.
Like China, the U.K. is also seeking to boost its data center capacity, but is facing stiffer opposition from the public.
Thousands protest against AI data centers in UK
In the U.K., the government’s desire to keep the country competitive in a global AI race dominated by China and the U.S. appears out of step with public mood, as thousands turned out last week to protest against the building of large-scale AI data centers.
According to the BBC, on November 14, thousands of people joined a protest in Edinburgh against the building of large-scale AI data centers.
The march came less than three weeks after new U.K. Chancellor John Healey used his October 28 Budget to launch the “Data Centre Covenant,” which argues that: “Countries that can host and power data centres will ensure their sovereignty, strengthen economic resilience, and be better placed to capture the benefits of the digital economy, and help shape the future of AI.”
Under the covenant, data centers that meet the tests can get faster planning and grid connections through a new Ministerial Control Centre, chaired by AI Minister Kanishka Narayan. The covenant also estimates that the U.K. will need 6 GW of AI data center capacity by 2030, roughly three times its current operational capacity of around 2 GW.
Scotland is central to that push, and on January 29 the U.K. government named North Lanarkshire an AI Growth Zone, clearing the way for 500MW of capacity from Scottish operator DataVita. Meanwhile, the local outlet Edinburgh Evening News recently reported that over 20 hyperscale proposals have been lodged in Scotland, ranging from 150MW to 1 GW.
During the November 14 protests against these new developments, the BBC reported that demonstrators carried placards and shouted slogans such as “AI data centres threaten humanity” and “AI data centres: costing the Earth.”
In response to the outcry, the Scottish government reportedly told the BBC it would publish new planning guidance on data centers before the end of the year, adding that it was “developing a strategic approach to maximizing the public benefit from data centers located in Scotland.”
Whether or not all the proposed data centers go ahead, there’s a clear desire from the top for the U.K. to be competitive in the AI field, which—despite scandals, rumored bubbles, and apocalyptic warnings—appears to have captured the imaginations of governments around the world, for better and worse.
FAQs:
What is the AI lawsuit about?
Four nonprofits claim Anthropic, Google, Meta, OpenAI, and SpaceX AI entities violate California’s Unfair Competition Law by developing AI that poses catastrophic risks.
What does the AI lawsuit want?
A ban on releasing models more capable than each company’s best system as of November 12, plus a ban on AI autonomously conducting AI research.
Why did Nvidia stock drop?
Reports linked CEO Jensen Huang to a suspected Chinese intelligence recruit. Yahoo Finance estimated an 18.29% fall, erasing about $913 billion in value.
Did Jensen Huang resign?
Yes. Huang resigned as Nvidia’s president, CEO, and board member on November 12.
How much is China spending on AI data centers?
Reuters identified over 23 billion yuan ($3.4 billion) in Chinese projects, many of which specify Nvidia H100, H200, or H800 chips.
Why are Scots protesting AI data centers?
Thousands marched in Edinburgh on November 14, citing environmental costs and risks to humanity, as over 20 hyperscale proposals target Scotland.
In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership—allowing it to keep data safe while also guaranteeing the immutability of data. Check out CoinGeek’s coverage on this emerging tech to learn more why Enterprise blockchain will be the backbone of AI.
Watch: AI and Big Data will be a long-term trend




