# Marc Benioff Tells AI Companies to Regulate Themselves or Face Lawsuits > Source: > Published: 2026-09-17 09:19:39+00:00 *Marc Benioff wants AI companies held to the same standard as a car maker whose brakes fail, and he says lawsuits, not new laws, might be what gets them there.* Marc Benioff picked an odd week to sound like the most cautious person in Silicon Valley. Nvidia's Jensen Huang had just told regulators to back off. JD Vance had called AI executives' pleas for oversight a "Trojan horse." Then the Salesforce CEO sat down with Fortune and said something almost nobody in his position was saying: regulate yourselves, or the courts will do it for you. In an interview Fortune published on September 16, 2026, timed to Salesforce's own Dreamforce conference in San Francisco, Benioff compared AI accountability to auto liability law. Car manufacturers get sued when a vehicle malfunctions and someone gets hurt. He argued AI companies should carry that same exposure, using existing product liability statutes rather than waiting on Congress to write anything new. He stopped short of calling for government regulation outright. That's a real distinction. It's the one that separates him from Anthropic's Dario Amodei, who has pushed for industry-wide standards and an actual slowdown. Benioff's version puts the burden on companies first, with the law as a backstop only if they fail to police themselves. He was specific about the model he wants to avoid repeating. "A lot of individuals got hurt through social media, we don't want that to happen in AI," Benioff told Fortune. It's a historical reference, not a hypothetical: he's pointing at a decade of underregulated platforms and asking the AI industry not to run the same experiment twice. [Gallup poll finds seven in ten Americans oppose AI data centers nearby](https://startupfortune.com/gallup-poll-finds-seven-in-ten-americans-oppose-ai-data-centers-nearby/) Four major polls this year, from Gallup to Fox News to Quinnipiac, all find that roughly two-thirds to seven in ten Americans oppose AI data centers being built in their own communities. The opposition cuts across party lines and centers on electricity costs, water use and noise, even as Trump dismisses the backlash and the industry commits... - [american public opinion on ai data centers](https://startupfortune.com/gallup-poll-finds-seven-in-ten-americans-oppose-ai-data-centers-nearby/) - [why americans oppose artificial intelligence data centers](https://startupfortune.com/gallup-poll-finds-seven-in-ten-americans-oppose-ai-data-centers-nearby/) ## Three answers on one stage Benioff wasn't just commenting on this debate at Dreamforce. He hosted it. He personally interviewed Amodei, OpenAI's Sam Altman and Nvidia's Huang on stage in front of more than 43,000 attendees, according to diginomica's account of the panel, and the three gave three different answers. Amodei stuck to his call for industry standards and a slower pace. Altman said he largely agrees regulation is needed but stopped short of endorsing a broad slowdown. Huang wanted none of it. "We don't need any new laws. We don't need new regulations," he told the crowd, urging companies to "run as fast as you can," according to TechCrunch's report on his remarks. That's the fault line Benioff is standing on. Not whether AI should be dangerous, which nobody argues for, but who pays when it is. Vance made the same fight overtly political days earlier. Speaking to reporters in Kansas on September 14, 2026, the vice president said AI companies asking for regulation "feels a bit strange," calling it "somewhat like a Trojan horse" aimed at locking out smaller competitors, according to CNN's reporting. He named Amodei, Altman, Google DeepMind's Demis Hassabis and Elon Musk as the executives pushing that line. Benioff wasn't on it. He's arguing for accountability through the courtroom rather than the White House, a different fight entirely, and one Vance's framing doesn't really touch. ## Not a one-off Benioff has been building toward this position for months, not just this week. That's the record. He first called AI chatbots "suicide coaches" in a Davos interview in January 2026, after OpenAI disclosed that more than a million ChatGPT users a week show signs of potential suicidal planning during conversations with the bot, a figure CNBC reported at the time. He's also raised Section 230, the law that shields platforms from liability for what users post, and questioned whether it should protect chatbots the way it protects a comment section. At Dreamforce, he floated one more idea beyond the lawsuit threat: a Fortune 500 style list ranking companies by ethical standards rather than revenue, with Apple singled out as the kind of company he'd put near the top for security practices. It's a small proposal next to the liability argument. But it points at the same instinct. Benioff doesn't want to wait for Washington to define what a responsible AI company looks like. He wants the market, and if that fails, the courts, to define it instead. **Also read:** [Kevin Warsh Says AI Hyperscalers Are Driving Up Your Borrowing Costs](https://startupfortune.com/kevin-warsh-says-ai-hyperscalers-are-driving-up-your-borrowing-costs/) • [King Charles Tells AI Chiefs at Scotland Summit Not to Lose Control](https://startupfortune.com/king-charles-tells-ai-chiefs-at-scotland-summit-not-to-lose-control/) • [House Passes Ratepayer Protection Act 417-3 to Shield Power Bills From Data Centers](https://startupfortune.com/house-passes-ratepayer-protection-act-417-3-to-shield-power-bills-from-data-centers/) [Startups Have a Narrow Window to Get Agentic AI Right Before Rivals Do](https://startupfortune.com/startups-have-a-narrow-window-to-get-agentic-ai-right-before-rivals-do/) Enterprises are racing to expand agentic AI in 2026, but skills, workflow redesign and trust are lagging far behind the technology itself. New data from CrewAI, Workera, Gartner and Section shows why startups that build all three now, rather than waiting, are the ones positioned to win. - [how to implement agentic AI before competitors](https://startupfortune.com/startups-have-a-narrow-window-to-get-agentic-ai-right-before-rivals-do/) - [enterprise agentic AI adoption strategy for startups](https://startupfortune.com/startups-have-a-narrow-window-to-get-agentic-ai-right-before-rivals-do/) *This article is posted in [AI News](https://startupfortune.com/category/ai/), check it out for more related stories.* ## Join the discussion [Open in the community →](https://startupfortune.com/community/) Almost there. Sign in and your reply posts straight away.