Leaked Microsoft Data Exposes Wide Gap in Employee AI Spending

Generative AI has become embedded in software development, research and administrative work at Microsoft, but model inference, premium subscriptions and computing resources can carry significant costs. Internal spending data offers a rare view into how unevenly the technology is being adopted across teams. It also underscores a broader shift among large technology companies toward measuring AI productivity gains against the expense of deploying and operating the tools at scale.

Recently leaked Microsoft data showed that median monthly spending on AI tools was about $300 per employee, while the highest monthly bill for a single employee or account reached $28,000. The wide gap suggests that a relatively small number of intensive users or teams can drive a disproportionate share of costs. The figures also point to tighter internal monitoring as Microsoft tracks resource consumption, identifies unusual spending and seeks to control rapidly growing AI-related expenses.