Accenture has limited centrally funded spending on Anthropic’s Claude to $29 per employee per month, requiring employees who exceed the allowance to identify project budgets to cover additional usage, according to a report by 404 Media.
The policy, which took effect on October 1, comes as enterprises face rising costs from employees using AI tools for coding and other intensive tasks. Some Accenture employees reportedly spent between $500 and $1,000 a month on Claude, raising concerns about how spending limits could affect access to AI tools and their use in development workflows.
The new policy applies to employees using Claude Enterprise, with additional usage costs charged to individual projects rather than covered by a central budget. Employees must identify a project through Accenture’s Work Breakdown Structure (WBS) to pay for usage beyond the monthly allowance. An internal announcement reviewed by 404 Media stated that “all costs are consumption-based”, explaining the shift towards project-level budgets for additional usage.
The restrictions have prompted some employees to explore higher usage tiers. According to 404 Media, Claude’s Premium plan includes $250 in token usage per week, and some employees have sought upgrades to avoid charging additional costs to their projects. An internal message reviewed by the publication also indicated that at least one employee had their Claude access revoked.
Accenture’s internal notice said licences would be reviewed periodically and could be reassigned based on business priorities, job requirements, usage patterns and workforce needs.
The policy highlights the challenge enterprises face in balancing AI adoption with the cost of running increasingly intensive workloads. Coding tools such as Claude Code can consume significant amounts of tokens during complex development tasks, making spending difficult to manage when usage scales across large teams.
The spending cap comes as businesses increasingly evaluate the costs associated with deploying AI tools across their workforces. While access to AI assistants can help employees automate tasks and accelerate software development, usage-based pricing can make expenses harder to predict, particularly for teams relying on these tools for extended coding sessions.
Accenture’s approach shifts some of the responsibility for AI spending from central budgets to individual projects, potentially requiring teams to weigh the costs of additional usage against their budgets and expected productivity gains.
This article was originally published on Mint – AI. Click below to read the complete article.