What an AI “slowdown” could look like

Anthropic researcher Hilbert Spaess publicly resigned and warned that the people building AI "earnestly believe it could kill us all by the end of the decade," as leading AI researchers and executives called over the weekend for new safeguards on the technology. Steve Bannon, Bernie Sanders, Republican Sen. Marsha Blackburn, Congressional Progressive Caucus chair Greg Casar, and the archbishop of San Francisco are set to speak at a "Pro-Human Assembly" in Washington on Tuesday, while President Donald Trump opposed the push on Truth Social, writing that "there is a SICK conspiracy going on against AI and Data Centers." Vox's Andrew Prokop outlined three possible federal approaches: banning AI systems that exceed human performance, empowering third-party auditors to review AI before public use, and holding AI companies accountable for harm.

This story appeared in Today, Explained, a daily newsletter that helps you understand the most compelling news and stories of the day. Subscribe here https://www.vox.com/today-explained-newsletter . What an AI “slowdown” could look like The leaders of several AI companies called for new safeguards over the weekend. Here are three things government could do. Caitlin Dewey https://www.vox.com/authors/caitlin-dewey is a senior writer and editor at Vox, where she helms the Today, Explained newsletter. If you’d asked me yesterday what Steve Bannon and Bernie Sanders had in common, I might have said “white hair.” But after a wild week https://www.vox.com/politics/502717/ai-freakout-mainstream-openai-anthropic-humans in which AI doomerism burst solidly into the mainstream, the right-wing provocateur and progressive senator are joining forces — with various religious leaders, lawmakers, and the actor Joseph Gordon-Levitt, no less — to express their deep concern about AI regulation https://www.axios.com/2026/09/14/coxon-bernie-bannon-unite-human-ai . Concern about cutting-edge AI models had been mounting for months as those models increasingly behaved in ways https://www.washingtonpost.com/technology/2026/09/11/ai-experts-warn-technology-is-learning-cheat-hack/ their makers didn’t expect. It exploded last week, when a researcher at the AI company Anthropic resigned publicly https://x.com/hilbertspaess/status/2097476196791709843 and warned that the people building AI “earnestly believe it could kill us all by the end of the decade.” Do you have questions about this week’s AI panic? Candidly: me too Submit your questions about AI safety and regulation through this form https://forms.gle/pBJhdGza9rWwrM7m9 , and I’ll hunt down the answers for an upcoming edition. It sounds like science fiction. It is the stuff of science fiction. But it’s not a fringe position https://www.nytimes.com/2026/09/12/technology/doomsday-discussions-ai-companies.html : Over the weekend, leading AI researchers and executives also warned that AI is developing dangerously fast and called for measures to rein it in. We’ll run through those options in just a minute. Now, a wide spectrum of lawmakers, activists, and pundits are also calling for stronger safeguards — Sanders and Bannon among them. The pair will speak at an event billed as a “Pro-Human Assembly” in Washington on Tuesday alongside Republican Sen. Marsha Blackburn, Congressional Progressive Caucus chair Greg Casar, and the archbishop of San Francisco, among others. But President Donald Trump is decidedly not on board https://www.npr.org/2026/09/13/nx-s1-5968078/trump-mike-johnson-ai-slowdown , and it’s hard to see how any substantive AI regulations could pass without him. “There is a SICK conspiracy going on against AI and Data Centers,” he wrote https://truthsocial.com/@realDonaldTrump/posts/117269745153543631 this morning on Truth Social, “and the only one happy about it is China.” Go a little deeper How could the federal government make AI safer, if it could ever muster the bipartisan political will to act? My colleague Andrew Prokop recently put this question to a bunch of AI safety advocates and legal experts https://www.vox.com/politics/502717/ai-freakout-mainstream-openai-anthropic-humans , and landed on three approaches: 1. Ban AI altogether. Federal law could hypothetically bar the development of AI systems that exceed human performance, or pause work on those systems until the government can stand up a new agency to monitor them. 2. Empower auditors. Some third-party authority — whether a government agency, a nonprofit technical organization, or an industry group — could begin reviewing AI systems before they go into public use. The Trump administration does a version of this already, but the process is voluntary and highly secretive. And it’s not clear if the officials reviewing these systems are super tech-savvy themselves. 3. Up the stakes. Congress could hold AI companies accountable for any harm that their systems or agents cause through changes to federal liability law. The threat of a huge, potentially bankrupting lawsuit could prompt those companies to be more cautious. All three approaches have precedent in federal regulation, so they wouldn’t be at least in theory too tricky to stand up. The real roadblock, as always, is political. One AI researcher told Vox he worries we’ll miss the opportunity to regulate AI while waiting for lawmakers to get on the same page. Even under threat of apocalypse, Washington, DC, moves far, far slower than Silicon Valley. One link for later ➨ Don’t skip leg day. Make that leg days . Experts recommend that healthy adults incorporate strength training into their schedules https://www.youtube.com/watch?v=Z0VJFcpVLm0 at least twice a week. It builds strength, confidence, and body awareness, and can help preserve muscle mass and bone density as you age. Before you go… - Did you know… that 93,000 wild horses still roam the Western US https://www.vox.com/future-perfect/501530/wild-horse-slaughter-blm-burros ? They’re descended from animals introduced by European colonists in the 1500s. - Today’s trivia: What pro hockey player went on to found a major restaurant chain with locations across Canada and the US? You can find this and other brain puzzles in Vox’s daily crossword https://link.vox.com/click/6a46cfd46352b082500e86fe/aHR0cHM6Ly93d3cudm94LmNvbS8yMTUyMzIxMi9jcm9zc3dvcmQtcHV6emxlcy1mcmVlLWRhaWx5LXByaW50YWJsZT91ZWlkPWM3Mzg0ZDJhYzRmZDQ0YzcyZDI4OGQ4N2YxZjM5NDlh/664389def568f1a4620aa3aeC7b08b10b . Look for the answer in tomorrow’s edition. - Yesterday’s trivia: Last week we asked you which country voted against restarting membership talks with the European Union this year. That would be Iceland, where control over the country’s fisheries https://www.politico.eu/article/iceland-referendum-question-marks-no-camp-successful/ has been a major barrier to EU accession.