Benioff Urges AI Companies to Embrace Hawaiian 'Kuleana' Concept of Responsibility

Marc Benioff stood at Salesforce's annual Dreamforce conference in San Francisco and made an argument that cut against the grain of how technology companies typically operate: they should be held responsible for the risks their products create before anyone gets hurt, not after.

The Salesforce CEO, speaking to Fortune, grounded this idea in a Hawaiian concept called kuleana—a word that carries the weight of personal responsibility and obligation. Benioff has spent the last few years deepening his connection to Hawaii, where he now lives. In 2024, he and his wife Lynne donated $150 million to two state hospitals, a gift described as the largest healthcare donation in Hawaii's history. That money was meant to expand access to care and shore up aging infrastructure. Now he was applying the same principle of accountability to the artificial intelligence industry.

"We know we have to hold companies responsible for their products and their technology before people are hurt," Benioff said. The statement was direct, almost simple—but it represented a departure from how most tech firms have approached emerging risks. The pattern has been familiar: build first, apologize later. Benioff was proposing something different. Companies developing AI systems, he argued, need to think through possible harms before those harms materialize, not scramble to explain them afterward.

He pointed out that the companies building these systems possess knowledge no one else does. "Only [tech] companies know what's going on in their lab," he said. That asymmetry of information creates an asymmetry of responsibility. The firms developing frontier AI models understand their capabilities and limitations in ways regulators and the public cannot. That understanding, Benioff suggested, should translate into action. "At some deep level, these companies must hold themselves responsible for their safety."

When asked about how such accountability might actually work, Benioff pointed to product liability law as one possible mechanism—a legal framework that already exists in other industries and could be adapted to hold AI companies answerable when their technology causes demonstrable harm. He stopped short of endorsing government regulation outright, declining to say whether lawmakers should step in to police the sector.

The timing of his remarks mattered. At the same conference, Benioff had interviewed OpenAI CEO Sam Altman, who referenced a July incident involving Hugging Face and autonomous AI agents as a cautionary tale. Altman suggested that companies need to ensure safety keeps pace with capability. Anthropic CEO Dario Amodei, appearing during Benioff's keynote, voiced support for slowing the development of frontier AI models. But not everyone agreed. Nvidia CEO Jensen Huang offered a counterpoint: speed and safety, he argued, are not mutually exclusive. Both can happen at the same time.

Benioff's invocation of kuleana—a concept rooted in Hawaiian culture that speaks to duty and stewardship—was an attempt to reframe the conversation. He was not calling for punishment or blame. He was calling for companies to internalize the weight of what they are building, to feel the responsibility as something personal and moral, not merely legal. Whether that appeal would move the industry remained an open question.