Companies are increasingly using artificial intelligence budgets to establish an informal workplace hierarchy, determining which employees and departments receive access to the most advanced and expensive tools based on their perceived value and potential return on investment, Bloomberg reports.
As AI costs rise, businesses are imposing spending limits or reserving premium access for revenue-generating teams, particularly engineering and research and development, while restricting access for departments such as finance. Apple, for example, recently reduced some engineers’ daily AI spending limits from $300 to $200, while employees outside engineering face a $100 cap. Accenture has withdrawn specialized Claude Enterprise licenses from some employees, and Goldman Sachs has restricted certain AI coding tools to engineers.
According to expense-management platform Ramp, median corporate AI spending per employee tripled from $5 to $15 monthly between September 2025 and September 2026, while spending among the top 1% of biggest AI spenders increased from approximately $2,500 to more than $9,500 per employee.
Some companies are tying additional AI funding to demonstrated productivity, work quality and measurable business results. At software company Sonar, executives track engineers’ output, quality and speed before approving larger budgets, while Cisco has granted unlimited AI access to an engineer whose work demonstrated exceptional value.
However, unequal access raises concerns about fairness, particularly when employees are evaluated on performance despite having different tools available. JLL plans to instruct managers not to penalize employees for having less access to AI tools during performance reviews.
