Artificial intelligence developers are expressing alarm that AI could destroy humanity in the next few years, and industry leaders are calling for a slowdown in its development. Congress, however, seems uncertain about what role it should play in reining in the onrushing technology.
‘Kill all humans!’
AI researchers believe that now is “crunch time for humanity,” said Jacob Coxon to Wired. (Coxon recently resigned from a short stint at Anthropic, following several years at OpenAI.) Workers at the big companies building artificial intelligence “earnestly believe that it could kill us all by the end of the decade,” he said on X.
Anthropic’s Evan Hubinger agrees. AI really could “kill all humans!” he said on X. And such warnings present AI companies with a “dilemma,” said Axios: “Slow down and risk falling behind, or press ahead and risk losing control.”
The Week
Escape your echo chamber. Get the facts behind the news, plus analysis from multiple perspectives.
SUBSCRIBE & SAVE
Sign up for The Week's Free Newsletters
Join more than 350,000 subscribers and keep yourself informed with a selection of The Week’s most interesting, enlightening and entertaining stories - plus daily puzzles.
Join more than 350,000 subscribers and keep yourself informed with a selection of The Week’s most interesting, enlightening and entertaining stories - plus daily puzzles.
Latest Videos FromThe Week
Coxon’s high-profile alarm is “part of a wave of warnings about unchecked AI systems,” said Ian Prasad Philbrick at Slate. The trend raises the question of why companies like Anthropic and OpenAI do not stop or slow their work on increasingly complex and capable models.
Some developers have “talked themselves into continuing to work” on AI, believing that their less-conscientious industry colleagues will otherwise have the playing field to themselves, said Philbrick. But that kind of thinking is a “one-way ratchet toward greater AI risk.” Still, the risks might be “overstated.” Voter anger over data centers may also “compel Congress to act” with rules and guardrails.
The warnings “merit a certain skepticism,” said The Washington Post in an editorial. (The Post is owned by Jeff Bezos, who cofounded the Prometheus AI lab in 2025.) The people working most closely with AI do not have “special knowledge of what machines that do not yet exist will someday become.” And they can’t say how “society will adapt to the challenge.”
But their warnings could spur Congress to overregulate the sector, which may curb the development of a promising new technology. “The dawn of the AI era cries out for humility and patience,” said the Post.
‘Monumental human stupidity’
Coxon’s and Hubinger’s concerns are “unpersuasive,” said Thomas Claburn at The Register. A “Skynet scenario” would “require monumental human stupidity” in the face of AI superintelligence. But new laws should be enacted to “put executives in jail when their models do harm.” If AI models really are that dangerous, society should “hold people accountable for the harm they cause.”
Congress “won’t lead the charge” on regulating AI, said Axios. Guardrails are needed, but “we have got to make sure that we don’t lose our edge in the global competition on this with China,” said House Speaker Mike Johnson on NBC’s “Meet the Press.”
President Donald Trump is also hesitant to act. Artificial intelligence is “going to be more good than bad,” he said to reporters, per The Associated Press.
