Benioff Tells AI Industry to Regulate Itself or Face Liability
Salesforce CEO Marc Benioff used Dreamforce to warn AI companies they must take responsibility for their products before people are harmed, comparing the moment to early social media failures and suggesting product liability law as a model for accountability.
Salesforce Chief Executive Marc Benioff used his company's annual Dreamforce conference in San Francisco to deliver a blunt message to the AI industry: companies that build powerful artificial intelligence systems must hold themselves accountable for the risks they create, or they will eventually face legal consequences.
Speaking in a walking interview with Fortune as he moved through the streets near Salesforce Tower, Benioff said the responsibility for AI safety rests primarily with the companies developing the technology. He compared the current moment to the early mistakes made by social media platforms, which he suggested failed to anticipate harm before it occurred.
«We know we have to hold companies responsible for their products and their technology before people are hurt,» Benioff said, invoking the Hawaiian concept of kuleana, or personal responsibility, as a guiding principle for corporate ethics. Benioff has long woven Hawaiian norms into Salesforce's corporate culture.
The remarks came during a Dreamforce week that has once again made the conference a focal point for debates over AI safety. Earlier in the day, Benioff interviewed OpenAI Chief Executive Sam Altman on stage at the Yerba Buena Center for the Arts. Altman described a July hack of Hugging Face by a swarm of rogue OpenAI agents as a terrifying wakeup call, and said companies should pace development so that safety stays ahead of new capabilities.
Anthropic Chief Executive Dario Amodei joined Benioff during his morning keynote at the Moscone Center and also advocated pacing frontier AI models. Nvidia Chief Executive Jensen Huang took a different view, arguing that speed and safety can exist simultaneously. Meta Chief Executive Mark Zuckerberg separately shrugged off concerns, writing that AI labs have a natural incentive to build safe systems.
Benioff declined to say whether governments should regulate AI companies directly, but he pointed to product liability law as a legal mechanism for accountability. He compared the idea to the way car manufacturers are held liable when a vehicle malfunctions. He also proposed a new Fortune 500-like list that would rank companies by their ethical standards, naming Apple as a firm he holds in high regard as a security standard-bearer.
«Only companies know what's going on in their lab,» Benioff said. «At some deep level, these companies must hold themselves responsible for their safety.»
At Salesforce, that philosophy has translated into wrapping AI models in a so-called trust layer designed to prevent misbehavior by operating within a highly constrained structure that also closely monitors agents. Benioff said neither Salesforce nor its clients have experienced Hugging Face-like episodes, noting that Salesforce is not itself creating the most powerful frontier models.
Benioff's relationship with the AI labs is complicated. Anthropic and OpenAI's autonomous systems threaten the core of Salesforce's product line, yet Salesforce also relies on their models to power parts of its flagship AI product, Agentforce. On Tuesday, Salesforce announced a new system allowing customers to access its tools without logging in, and unveiled a new reasoning model for Agentforce built with Nvidia.
The CEO's hometown standing has also been tested. Last year he apologized after calling on President Trump to deploy the National Guard to San Francisco during Dreamforce, citing a shortage of local police officers. Angel investor Ron Conway, a longtime friend, resigned from the Salesforce Foundation's board after the comments, which disappointed residents who had known Benioff as a progressive, Democratic-supporting CEO before he more recently embraced Trump. Benioff has said he is an independent.
On Tuesday, as protesters gathered near the Moscone Center over issues including Salesforce's work with U.S. Immigration and Customs Enforcement, Benioff struck a lighter tone. He said the company still hires hundreds of off-duty officers during the conference, a practice that has drawn scrutiny in the past.
Benioff also said he had not spoken recently with President Trump about AI safety. Over the weekend, Trump called AI doomsday scenarios exaggerated and blamed «negative forces» for the warnings. Benioff waved off concerns that AI could wipe out humanity, a topic that has consumed Silicon Valley since Anthropic researcher Jacob Coxon quit over such worries.
