Salesforce CEO blames Microsoft for not investing in OpenAI

Salesforce CEO Marc Benioff wanted a stake in OpenAI long before most of the tech industry was chasing one. Sam Altman turned him down, and Microsoft was the reason. Speaking on Alex Heath's Sources podcast, in an episode published on Tuesday, Benioff said OpenAI's ties to Microsoft, a direct rival to Salesforce in enterprise software, kept the door shut. So Salesforce went to OpenAI's biggest rival instead. That fallback, Anthropic, is now valued at $965 billion and preparing for a public listing. Benioff believes the bet could earn Salesforce "not just billions but maybe tens of billions of dollars." Microsoft's 27% OpenAI stake kept Salesforce off the cap table "I knew Sam Altman very well," Benioff said on the Sources podcast. "I was like, 'Hey, I'd really like to be investing now in the company.'" According to Benioff, Altman replied that the Microsoft relationship meant Salesforce was not allowed to invest. On paper, the block made sense. Microsoft owns roughly 27% of OpenAI, uses its models in products that compete with Salesforce's, and sells OpenAI access through Azure. The two companies have reworked their deal twice since October 2025. The latest revision, in April, made Microsoft's licence to OpenAI technology non-exclusive and allowed OpenAI to serve customers on other clouds. Benioff did not claim foresight. "We felt like, whoa, we can't go here, we're going to have to go everywhere else," he told Heath, adding that it was "not because we were so smart." "Everywhere else" meant a wide spread of AI startups, including Hugging Face. Salesforce's $50 million Anthropic investment from 2023 is now worth around $5 billion Salesforce first joined an Anthropic funding round in early 2023 with about $50 million and kept investing in later rounds. By June this year, that stake was reportedly worth around $5 billion, a figure calculated after Anthropic's $65 billion raise. Salesforce has never disclosed exactly how much of the Claude maker it owns. An Anthropic IPO would put a public price on that holding and eventually give Salesforce a way to sell shares. Anthropic has confidentially filed draft paperwork with regulators, with Morgan Stanley and Goldman Sachs reportedly leading the offering. Altman, meanwhile, has said OpenAI will not go public in 2026. Anthropic's Claude now runs Salesforce's Slack and Agentforce "It's because we have been pushed so close with Anthropic," Benioff said on the podcast, explaining why he expects such large returns. The investment has become a working partnership. Claude is the default model across Slack and runs Slackbot, which Salesforce says has saved its employees 8.1 million hours of work on an annualised basis. In August, the two companies launched Claudeforce, which brings Salesforce data into Claude through a plugin with 37 prebuilt sales skills, from meeting prep to pipeline reviews. Claude also serves as a reasoning model inside Agentforce. Money moves the other way too. In May, Benioff said Salesforce would likely spend $300 million on Anthropic tokens this year, much of it on coding. Benioff lost out on OpenAI but never cut ties with it The Anthropic tie-up has not made Salesforce a one-model company. Both Dario Amodei and Altman appeared at Dreamforce 2026 in San Francisco last week, and Salesforce is adding OpenAI capabilities to Missionforce, its AI agent platform for government customers. Benioff used the same stage to call the so-called SaaSpocalypse "crazy nonsense" and to forecast that Salesforce would cross $50 billion in annual revenue next year. Three years after Altman said no, Salesforce holds an early stake in one of the world's most valuable AI companies and still has OpenAI in the room. For a company that was shut out, it ended up with a seat at both tables.