Marc Benioff, the Chief Executive of Salesforce, has stepped forward as the latest prominent tech leader to call for greater accountability within the artificial intelligence sector, a plea that comes as increasingly powerful AI models gain widespread adoption. His remarks arrive amidst a simmering debate over AI safety, which was recently intensified by an essay from Anthropic's CEO, Dario Amodei, over the weekend advocating for frontier AI labs to decelerate model development in order to buy more time for safety protocols to catch up. Benioff's comments add a significant voice to this growing chorus of concern.

Known for his long-standing advocacy that tech companies must also serve the public good, Benioff has a history of vocal criticism toward social media firms and their societal consequences. In a memorable 2018 statement, he famously compared Facebook to a new version of cigarettes and pushed for stronger government oversight of the industry. Now, he is cautioning AI enterprises against repeating what he views as past mistakes.

Speaking on the sidelines of Salesforce's annual Dreamforce gathering in San Francisco, Benioff expressed his concerns directly. "Social media has already inflicted harm on many businesses and individuals," he stated. "We do not want to see a repeat of that scenario in the AI landscape."

His warning carries particular weight given Salesforce's deep connections with the vanguard of the AI movement. Just last month, the company unveiled Claudeforce, a new offering that includes built-in plugins allowing clients to utilize Anthropic's Claude model to access customer data stored within Salesforce. This integration enables tasks such as drafting emails or updating records. On Tuesday, Amodei shared the stage with Benioff during the latter's keynote address. "I think this is the way to lead the industry forward, to set an example," Amodei remarked, "showing everyone that there is always room for improvement."

Benioff, however, stopped short of directly criticizing specific AI companies or endorsing calls to halt AI research and development. Instead, he advocated for a principled approach, arguing that the companies creating this technology have a duty to weigh its broader implications as they forge ahead. "We want AI to handle issues properly, to carry its responsibility, to regulate its own behavior, and to adhere to ethical principles," Benioff explained. "The core of ethics is accountability." He further added, "We genuinely care about our peers in San Francisco and around the world, hoping to say: 'We want you to uphold ethical standards in your actions, your treatment of the community, and your engagement with the world.' After all, this is a technology of immense power."