Salesforce, Uber shift budget to AI

- Marc Benioff said on May 16 Salesforce could spend about $300 million on Anthropic tokens in 2026, extending an AI-led slowdown in engineering hiring. - Uber had used its full 2026 AI budget by mid-April after Claude Code adoption spread internally, according to reports citing CTO Praveen Neppalli Naga. - Salesforce reports fiscal first-quarter results on May 28, when investors can look for updated comments from Benioff and finance executives.

Salesforce and Uber are putting unusually concrete numbers on a change many companies have described more vaguely: money that might once have funded hiring is being redirected to AI usage. Salesforce CEO Marc Benioff said on the All-In podcast published May 16 that the company could spend about $300 million on Anthropic tokens in 2026, much of it on coding. Separate reports on Uber said the ride-hailing company had exhausted its 2026 AI budget within four months as engineers adopted Anthropic’s Claude Code. UBS Global Research said on May 13 that artificial intelligence accounted for 26% of announced corporate layoffs in the latest month, citing Challenger, Gray & Christmas data. ### What exactly did Benioff say Salesforce would buy? Marc Benioff said Salesforce expects to spend roughly $300 million on Anthropic tokens, with most of that tied to coding work and software development. Reports on the podcast remarks said Benioff described coding agents as “awesome” and said AI made work cheaper and faster to produce. (finance.yahoo.com) The spending plan follows Benioff’s earlier comments that Salesforce would not add more software engineers in 2025 because AI had lifted productivity. Reports from 2025 and 2026 said he linked that hiring stance to gains from AI tools inside the company. ### Why do Anthropic “tokens” matter more than a software license count? Anthropic tokens are the metered units companies buy when employees or software systems use large language models through APIs and coding tools. (finance.yahoo.com) Salesforce’s projected $300 million outlay matters because it frames AI as an operating expense tied to consumption rather than a fixed software seat count, according to Benioff’s description of token purchases for coding-related work. (salesforceben.com) For companies, that means AI costs can rise with usage. Forbes reported on May 17 that Uber burned through its annual AI budget in four months on Claude Code, a sign that internal adoption can move faster than finance plans built around older software-pricing assumptions. ### What happened inside Uber? Uber had exhausted its 2026 AI budget by April after engineers rapidly adopted Claude Code, according to reports citing Chief Technology Officer Praveen Neppalli Naga. (economictimes.indiatimes.com) One report said Uber began a controlled rollout in December 2025, saw usage accelerate by February 2026 and was “back to the drawing board” by mid-April. (forbes.com) The company’s broader research-and-development spending remains much larger. Reports citing Uber’s finances said the company spends about $3.4 billion annually on R&D, but the AI budget overrun still stood out because it arrived so early in the year and was tied to developer tool adoption rather than a one-time infrastructure buildout. ### How much of the layoff data is being linked to AI? (finance.yahoo.com) UBS Global Research said on May 13 that 26% of announced corporate layoffs in the latest month were attributed to AI, based on Challenger, Gray & Christmas job-cut data. Challenger’s April report said AI accounted for roughly 16% of all 2026 job-cut plans year to date. Andy Challenger of Challenger, Gray & Christmas told Inc. that even when AI is not directly replacing a worker, “the money for those roles is,” as companies reallocate spending. (finance.yahoo.com) CBS News reported that economists are still debating how much of the shift reflects direct displacement versus broader restructuring. ### Is this a hiring freeze story or a spending story? Salesforce provides one example of both. (finance.yahoo.com) Benioff’s comments paired a hiring slowdown in engineering with a large planned token bill for AI coding tools. Uber provides another example from the spending side, where usage outpaced budget assumptions before the year was half over. UBS and Challenger data do not prove that every dollar of AI spend replaces a worker. (inc.com) But the figures show companies are now reporting AI in staffing decisions and, in some cases, disclosing material operating costs tied to model usage and coding assistants. ### What should investors and employees watch next? May 28 is Salesforce’s scheduled date for fiscal first-quarter results, when Benioff and finance executives may give more detail on AI-related spending, hiring and margins. (salesforceben.com) Uber’s next public markers are likely to come through earnings commentary, product disclosures or executive remarks about internal developer-tool costs and budget controls. (finance.yahoo.com 1) (finance.yahoo.com 2)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.