Salesforce’s Strategic Investment in AI Tokens
Marc Benioff, the CEO of Salesforce, has announced that his company anticipates an expenditure of $300 million on Anthropic tokens within the current year, primarily earmarked for coding endeavors.
This declaration came during the All-In podcast on Friday, where Benioff lauded AI coding agents, characterizing Anthropic as “awesome.”
He asserted that this financial commitment would render the developmental processes at Salesforce markedly more cost-effective.
In the realm of artificial intelligence (AI) and Large Language Models (LLMs), tokens serve as the foundational elements of text.
Essentially, when a prompt is entered into an AI framework, it disassembles the input into tokens for processing, subsequently generating the resultant output.
Each LLM, including ChatGPT and Notebook, possesses a specific token cap that determines the volume of data it can efficiently interpret and produce at one time. AI enterprises levy charges based on token usage, often around $2 for every million tokens.
Shift Towards AI-Driven Coding
Benioff’s recent plans, combined with his earlier statement regarding the cessation of engineer hiring in 2025 due to productivity enhancements through AI, indicate a significant pivot—namely, the automation of code writing.
The Salesforce leader maintains a strong belief that AI is not extinguishing entry-level opportunities but is rather reshaping them. Just last month, he proclaimed that the firm would recruit 1,000 new graduates “to harness the AI exponential.”
Over the past two years, however, Benioff has traversed a complex landscape, issuing announcements regarding layoffs, pausing hiring initiatives, and simultaneously onboarding recent graduates.
In February 2026, Salesforce reportedly eliminated nearly 1,000 positions within marketing and product management, attributing the reduction to AI automation.
By September 2025, a further announcement revealed the reduction of approximately 4,000 customer support roles, again as a direct consequence of AI advancements.
In December 2024, Benioff declared a freeze on hiring new software engineers, citing a remarkable 30% productivity increase facilitated by their in-house autonomous AI system, Agentforce, alongside other AI technologies.
Concurrently, he explicitly outlined a recruitment drive for 1,000 to 2,000 sales representatives, emphasizing the necessity of human presence to effectively convey and demonstrate intricate enterprise AI solutions to clients.
Reflecting on the evolving nature of leadership, Benioff remarked, “I’m probably the last CEO that you’ll meet who only managed humans.
All of us going forward are going to manage humans and agents together,” coining this transformation as the ‘digital labor revolution.’
An investment of $300 million in annual tokens from a single client would position Salesforce as one of Anthropic’s foremost commercial partnerships, although neither entity has validated this figure in their official communications.
Source link: Livemint.com.
