Salesforce Anthropic token spend signals AI as core cost
Salesforce Anthropic token spend is moving from a headline number to something more revealing: a sign that AI is becoming a core operating cost inside major software companies. Marc Benioff said on the podcast published Friday that Salesforce expects to spend $300 million on Anthropic tokens in 2026, with most of that bill tied to coding. That detail matters because it shifts the story away from flashy demos and toward day-to-day economics. Benioff is not talking about a small pilot. Instead, he is describing a level of AI usage that suggests coding agents are being treated as infrastructure. He also tied that spending to a broader product push. Salesforce is working on technology to make coding easier inside Slack, extending the companys bet that workplace collaboration and AI agents will increasingly blur together. Salesforce Anthropic token spend points to a bigger AI cost shift Benioffs projection, shared on the podcast, puts a price tag on how deeply Salesforce appears willing to embed Claude into its internal workflows. He said most of the expected $300 million token spend will go toward coding, a striking signal for an enterprise software company whose own products are built by large engineering teams. The implication is simple: Salesforce believes AI