Bill Gates spent decades as one of technology's most reliable optimists. This week, on NBC's Meet the Press, he said AI is powerful enough to help cause a billion deaths. Bernie Sanders, meanwhile, is calling for an outright ban on what he terms artificial superintelligence. Two Republican governors are pushing state-level AI regulation despite their president's dismissal of the entire concern as a hoax. None of this sorts neatly along party lines, and that's the actual story.
Gates's Escalation Has a Clear Arc
This didn't come out of nowhere. In July 2023, Gates wrote that the best reason to believe AI risk is manageable is that humanity has managed comparable risks before. By January 2026, he'd narrowed his concern considerably: An even greater risk than a natural pandemic, he said, is a non-government group using open-source AI to engineer one. In August, a memo he published argued AI would end up either the greatest equalizer in history or the worst source of injustice, and that on the current course, a net negative outcome looked more likely than not.
This week's comments are the sharpest yet. Gates told Meet the Press that AI is certainly powerful enough to drive events that cause a billion deaths, and that there has never been a weapon as powerful as the combination of people with bad intentions and today's AI tools. His framing is worth noting precisely: He isn't describing a machine that spontaneously turns on its creators. He's describing a tool that makes a person who already wants to cause mass harm dramatically more capable of doing it. He also said a kill switch, the industry's most commonly cited safeguard, won't be sufficient on its own, and that getting the world to agree on real safeguards will be harder than anything the Cold War required.
Gates has said he's willing to make this case to President Trump directly, and has publicly disagreed with Trump's characterization of AI safety warnings as a hoax.
The Left Isn't Unified Either
Sanders's push for a ban on artificial superintelligence puts him at the sharper end of Democratic opinion, not the center of it. Reporting from Wired describes the Democratic coalition as genuinely split over AI doom: most Democrats side with Sanders's underlying concern, but a meaningful faction of progressives argue that the AI companies themselves are the ones promoting the most extreme extinction-level scenarios, and that doing so conveniently distracts from harms that are already measurable today: data-center power and water consumption, job displacement, and algorithmic bias among them.
Two Genuinely Different Arguments Against Regulation
It's worth separating two distinct skeptical positions that both land against new AI regulation, because they don't agree with each other any more than the pro-regulation camp does. The Washington Post's editorial board has argued that AI models are still products built by corporations, and that existing legal liability, the threat of ruinous lawsuits and damages, is already a sufficient mechanism to make companies fear the consequences of reckless behavior, without needing to regulate an unknowable future in advance. A separate, more libertarian argument holds that government control over which AI products may be developed would primarily entrench today's largest players, who would help write the rules they'd then be best positioned to navigate.
Both arguments conclude that new regulation is the wrong tool. They disagree on almost everything else, including whether the underlying risk is real.
What Actually Moves From Here
Congress has passed no broad federal AI legislation, and Senate leadership has been noncommittal about whether that changes before year's end. State-level action is where the real near-term movement is happening: California Governor Gavin Newsom signed an executive order this month fast-tracking AI oversight, and DeSantis's and Cox's state-level pushes are further evidence that federal gridlock isn't stopping regulation, just relocating it. A meeting between Trump, Speaker Mike Johnson, and AI industry executives has reportedly been on the calendar, though its actual focus remains unclear.
Sources: AI Pulse · Big Picture · workplaceai.ai. Gates's Meet the Press comments and warning history: NBC News, The Washington Post, Al Jazeera, Axios, and Forbes, September 25-28, 2026. The realignment across both parties: The Atlantic, Politico, and Wired, as synthesized by The Week, September 28, 2026. Hinton's catastrophe estimate: Axios. Newsom's executive order: Deadline. Every quote and figure above is attributed to its original reporting; none is a WorkplaceAI study, and WorkplaceAI takes no position on the underlying policy debate.