Vermont Senator Bernie Sanders on Monday warned the heads of OpenAI, Anthropic and Meta about the dangers of unregulated artificial intelligence, urging them to take action before it is "too late".
Mr Sanders, an Independent, has regularly warned of the potential for labour disruption and the environmental impact of unregulated AI.
“Almost every day, there is a new story about how your companies are losing control of the AI technology you are developing, with potentially cataclysmic results,” he wrote. “It is not too late to avoid disaster … stop building machines that humans cannot control.”
He was referring to incidents in which AI models owned by OpenAI, Anthropic and Meta went rogue, with the big-tech giants unaware they had managed to compromise the digital infrastructure of various organisations and companies. Mr Sanders said the events should serve as a wake-up call.
The Future of Life Institute, a non-profit group based in the US, says time is running out to maintain “meaningful human control” over AI.
“The companies admit that they cannot control their systems effectively, yet keep making them more dangerous anyway,” said Anthony Aguirre, president and chief executive of the institute.
“They must stop the creation of these superhuman, autonomous AI systems and redirect AI development towards controllable and pro-human AI tools.”
Cybersecurity experts have long warned of AI going rogue and the spate of incidents has raised concerns around the world.
Mr Aguirre pointed out that for all the consternation among elected officials, US Congress has taken no action to regulate AI. Mr Sanders' letter said he would be encouraging colleagues in Congress to take up the matter of supervising firms and enforcing accountability.
Meta chief executive Mark Zuckerberg released his own letter on Monday, offering different views to those of Mr Sanders.
The tycoon in charge of Facebook, Instagram, Threads and WhatsApp outlined his vision for AI, saying his company's levels of investment and research would benefit all. “The best and most realistic path to building a positive AI future is by delivering superintelligence to everyone,” he said.
He also lamented public opposition to the technology. “It is surprising that the discourse from many developing AI is so filled with doom,” Mr Zuckerberg wrote.
“I do not understand why anyone who believes that AI will eliminate most jobs and much of humanity's relevance would rush to build that future.”
Mr Zuckerberg's letter comes days after Meta was fined $567 million by a New Mexico court for failing to protect teenagers and children who use Facebook and Instagram.
He said the push from governments to slow down AI developments was ill-conceived.
“Government policy is necessary to ensure a positive future and mitigating the risks,” Mr Zuckerberg wrote. “But it is also important for national security that the American AI industry continues to lead.”
Attempts by the White House to quietly apply more pressure on AI firms seems to have had limited impact.
Those moves by US President Donald Trump's administration have been critiqued by both proponents of more regulation who say the White House has been too secretive about what they are doing and those who say they want it to recommit to a hands-off approach.
“Secrecy invites abuse,” said Neil Chilson, former chief technologist for the Federal Trade Commission and current head of AI policy at the Abundance Institute.
“The rules will shift with each new administration and they will weight national security over economic growth, even though lasting security depends on a strong economy.”
For now, however, the public's worries about AI seem to be solidifying in many parts of the US.
Even in technology-friendly states such as California, where OpenAI, Anthropic and Meta are based, there is an increasing push for more accountability.
Democratic Governor Gavin Newsom on Monday revealed a new programme to protect California's critical infrastructure, with hopes it will improve the state's preparedness for a potential surge in “AI cybersecurity incidents”.
It comes months after Mr Newsom signed an executive order designed to “ensure companies protect people's rights” amid the AI boom.
