Four non-profit organisations have sued five leading US technology companies, asking for the development and use of new artificial intelligence (AI) models to be temporarily stopped. Through the court, they want the companies’ training and release of AI models suspended. The plaintiffs say the development and use of these technologies should resume only once the companies can show that human control is assured and the necessary safeguards are in place.
The coalition of groups working on digital safety and consumer rights filed the case in San Francisco Superior Court on Thursday. The plaintiffs are the Irreplaceable Education Fund, Public Citizen, the Heat Initiative and the Alliance for Secure AI. The case comes after the Democrats’ large win in the US midterm elections.
Many analysts believe the election result could change the regulatory framework for Silicon Valley, and that oversight of the technology sector could even become stricter.
Five technology companies are named as defendants: Anthropic, Google, Meta, OpenAI and SpaceX.
The plaintiffs allege that the companies are failing to prevent “catastrophic harm” caused by the AI technology they build. The groups want development and use of these technologies stopped until safeguards are assured.
The case cites documented incidents of cyberattacks carried out using AI and of other harmful behaviour by the technology. It also points to serious warnings issued by the companies themselves. Those warnings say that uncontrolled AI could slip out of human control and become a threat to humanity’s existence.
The plaintiffs have asked for an immediate emergency injunction on the big technology companies’ next-generation AI models. They say the companies should be able to resume work only when they can prove the technology is under human control.
The complaint says: “No other industry is allowed to operate on the assumption that building unsafe products is inevitable, especially products that could kill all of us. The plaintiffs have brought this case to stop the defendants’ unlawful path towards catastrophe.”
The plaintiffs allege that by putting public health at risk and disrupting the use of property such as computer systems, the companies have broken California’s unfair competition law. That law bans practices such as deceptive advertising, anti-competitive conduct in the market and luring buyers with one product in order to sell another.
The groups also allege that fast-advancing AI models are creating new crises almost every day, and that dealing with them has forced them to divert resources set aside for their regular work.
How a halt to AI development would work
If AI development is halted temporarily by court order, turning that into a permanent injunction raises several difficult questions. How the injunction would be enforced, how compliance would be checked, and what legal conditions would have to be met for it to be lifted are all still unclear.
One of the four plaintiffs, the Irreplaceable Education Fund, is a New York-based non-profit that works to make sure technology is used safely in the public interest. It told Al Jazeera that the main aim of its demand is to establish government control over AI development.
Its founder and executive director, Phil Aroneanu, said: “To genuinely pause development, you have to specify which activities must stop, how compliance will be verified and what the consequences of breaking the rules will be.”
He added: “That requires government inspection and the power to stop prohibited work. A ban on releasing AI models to the public alone will not be enough. Large computing centres could also be brought under monitoring to enforce the ban.”
On the legal safeguards AI development should have, the organisation has proposed several layers of oversight. In its view, decisions on AI development must be taken carefully and on the basis of public opinion.
The Irreplaceable Education Fund said several proposals could work. They include creating independent evaluators and a regulator at federal level, holding technology companies and their products to strict liability, and setting up an independent body to oversee AI.
The organisation says federal law should set minimum safeguards for AI safety, while states keep the power to adopt and enforce stricter measures.
Competition with China and AI safety
Although the case was filed in a California court, the plaintiffs argue that a pause on AI development in the United States would not hand an advantage to rivals such as China. For that, they say, Washington must lead on global rules and open talks with Beijing.
US technology companies have said repeatedly that the country is moving at full speed to outpace China in AI. President Donald Trump has taken the same position.
In a statement to Al Jazeera, the Irreplaceable Education Fund said: “We believe bilateral agreements and an international treaty will be needed to reduce the negative effects of AI and prevent catastrophe.”
The statement added: “If both sides’ progress can be slowed at roughly the same rate and the risk of losing control of AI is reduced, neither the United States nor China will lose much in the AI race. Instead, all of us will be safer.”
That argument runs against Trump’s long-held position. Trump, who opposes slowing AI development, fears that if the United States shows restraint on its own, Beijing will get the chance to set the direction of global AI development.
Trump, who backs the rapid expansion of AI, has said before that staying ahead of China in AI is essential. In his words: “Whoever wins AI, wins.”
The plaintiffs say that despite years of warnings about AI, the US Congress has yet to pass any law to regulate the technology.
One of the plaintiffs said: “As long as the companies keep racing to build superintelligent AI, more people and more critical infrastructure will be put at risk.”
The plaintiffs hope the Democratic lawmakers about to take office will respond to their demands. The Irreplaceable Education Fund said: “The new Congress’s first act should be to pause new AI development. But we cannot wait for lawmakers to act. The courts must now provide a remedy.”
বাংলায় মূল প্রতিবেদন পড়ুন · Read the original Bengali report