Salesforce's Benioff warns of AI 'false prophets' while promising true profits by 2030

CRM messiah preaches data discipline while rivals chase LLM miracles

Salesforce CEO Marc Benioff has warned investors to beware "false prophets" peddling AI salvation, as the CRM giant bets on its "agentic enterprise" vision to drive annual revenue past $60 billion by 2030.

The topline prediction came as the SaaS biz reported Q2 [PDF] financial reaults for the three months ended July 31, with revenue up 10 percent year-on-year to $10.24 billion and net profit jumping 32 percent to $1.89 billion.

Cloud and AI subscriptions were the standouts, with annual recurring revenue topping $1.2 billion – up 120 percent on last year.

Salesforce also handed $2.6 billion to shareholders through buybacks and dividends.

For Q3, sales are forecast to be between $10.24 billion and $10.29 billion, up eight to nine percent. Annual revenue is projected to smash through $60 billion by 2030, fueled by the corporation's growing stable of AI and data-driven services.

"We're confident we can deliver more than $60 billion in revenue by 2030," Benioff told analysts, calling the goal "a new chapter for Salesforce" as it morphs from cloud CRM vendor to what he described as a "digital labor platform." The target excludes the pending $8 billion acquisition of data integration specialist Informatica, which will deepen Salesforce's data governance stack and feed its AI pipeline.

Benioff reserved his sharpest words for the hype surrounding generative AI. "There's a lot of folks trying to be prophetic," he said during the investor Q&A. "Some of them are prophets and some of them are false prophets. It's going to be up to you to separate the wheat from the chaff."

His warning was aimed at rivals rushing to bolt large language models onto enterprise software. Salesforce insists its Agentforce system – the core of what executives call the "agentic enterprise" – connects those models directly to trusted corporate data and workflows instead of letting them drift unmoored in the cloud.

Chief product officer Steve Fisher told investors that Salesforce spent the past four years re-architecting its core platform to unify data, analytics, and automation around what is now branded Data 360. That, he said, allows companies to create "agents" able to act across sales, service, and marketing systems rather than just generate text.

"The LLM by itself isn't particularly useful for business," Fisher said. "It needs to be connected to the data, to the systems, so it can take action... It's about the humans and the agents working together."

Those agents are already being used inside Salesforce itself. The company's internal sales bots have made more than 100,000 calls to prospective customers, Benioff said, claiming they've helped close "hundreds of deals."

Chief revenue officer Miguel Milano added that more than 10,000 customers are now paying for Agentforce and Data 360 services, and that uptake has been strongest among smaller firms. "Every customer wants to become an agentic enterprise," he said. "They want to grow the top line, drive productivity, reduce cost, and empower employees. Now they know AI is going to enable that."

Benioff noted that government customers remain relatively slow adopters: "The government will, of course, come kind of at the end of the technology-adoption curve."

For now, Salesforce's AI narrative is helping offset slowing growth in its traditional CRM products. About 40 percent of bookings for Agentforce and Data 360 in the last quarter came from existing customers expanding usage rather than new sign-ups, a pattern Benioff framed as proof that the company's longstanding base is buying into its AI overhaul.

Still, the scale of the $60 billion promise leaves little room for error, and Salesforce must show that its "agentic" concept can generate sustainable margins rather than one-off excitement.

Benioff sounded convinced the bet will pay off. "We're not having to take back a lot of the things we've said over the last three years," he told investors. "We've been mostly on point and accurate in predicting the future… we've put A and B and C, and now we're going to deliver D." 

Here's hoping "D" doesn't stand for "Disappointment." ®