Microsoft Employee Spent $28,000 on AI in 28 Days as Internal Spreadsheet Exposes Usage Patterns

Microsoft AI Spending Reaches Unprecedented Levels

Microsoft employees are burning through AI budgets at rates that have caught the company's attention. An internal spreadsheet where staff voluntarily share compensation data now includes a column tracking monthly AI usage, revealing a median monthly AI spending of approximately $300 over 28 days across roughly 350 entries1

1

. The highest single figure recorded was a staggering $28,000, spent by an employee in the Customer and Partner Solutions organization2

2

.

This internal spreadsheet, which Microsoft employees use to compare salaries, bonuses, and stock awards, added the AI usage column in 2026. The data shows massive variation across teams. CoreAI reported a median of $975, Security came in at $526, and Microsoft AI at $490, while the Customer and Partner Solutions group showed the lowest median at $134 despite producing that $28,000 outlier1

1

. For context, a Microsoft 365 Copilot seat costs $30 a month, meaning the median Microsoft employee is spending ten times that amount on AI tool spending1

1

.

Excessive AI Usage Shows No Career Benefits

Despite the substantial AI costs, Business Insider found no meaningful correlation between how much AI someone used and their bonus rating, merit increase, or chance of promotion1

1

. This finding matters because it challenges assumptions about AI adoption driving career advancement. The data is self-reported, voluntary, and covers around 350 people out of Microsoft's 223,000 employees, but it reveals a troubling pattern where hours saved by AI are mostly thrown away rather than redirected to productive work1

1

.

Some departments saw extreme variations in token consumption. Within CoreAI, workers spent up to $16,000, while Cloud and AI saw figures reaching $15,000, and Security hit $10,0002

2

. These numbers emerged after years of Microsoft encouraging employees to embrace AI tools aggressively.

Company Cracks Down on Tokenmaxxing Trend

Microsoft leadership has begun tightening internal AI usage policies after observing what insiders call "tokenmaxxing"—a trend where users deliberately consume large amounts of AI tokens to become visible on tracking dashboards2

2

. In early August, Jay Parikh, executive vice president of CoreAI, sent employees a memo addressing the issue directly. "Tokenmaxxing is not what we are optimizing for," Parikh wrote. "I want all of us focused on maximizing outcomes that move the needle for our customers and our business"3

3

.

Parikh announced that Microsoft would manage token spend with the same discipline applied to every other critical resource and switch to OpenAI's GPT-5.6 Sol as its default for internal use—a less token-hungry model designed to reduce excessive AI usage3

3

. The company is now monitoring token consumption value more closely to track individuals who use large amounts of tokens2

2

.

Privacy and Regulatory Implications

The monitoring tools Microsoft uses to track AI usage raise questions in regions with strong worker protections. In Germany, works councils have co-determination rights over technical devices capable of monitoring behavior or performance1

1

. The Federal Labour Court has held since 1975 that a system only needs to be suitable for monitoring, not designed for it, which means even spreadsheets and calendars have triggered this rule1

1

. Any per-employee tracking tool would need a works agreement before reaching a German desk.

This trend isn't unique to Microsoft. Amazon previously held internal leaderboards ranking employees by AI usage, only to shut them down once the bills arrived3

3

. The industry is reckoning with the mindless enthusiasm for AI it created, as companies balance innovation with fiscal responsibility and meaningful AI outcomes.