Microsoft Tells Staff to Stop Tokenmaxxing AI Usage
Microsoft is formally curbing the tokenmaxxing culture by introducing budget targets and switching to a cheaper model, signaling a shift from unlimited AI use to disciplined spending.
Reporting from 1 source: GIGAZINE.
Microsoft is pushing back against 'tokenmaxxing,' a culture where employees maximize AI token use as proof of productivity. An internal email from Executive Vice President Jay Parikh obtained by 404 Media tells staff to avoid consuming excessive AI tokens and to focus on meaningful outcomes. Microsoft will update internal guidelines, set AI token budget targets per department from July 2026, and shift to a lower-cost OpenAI GPT-5.6 model.
Microsoft is reining in the practice of 'tokenmaxxing,' where employees burn through AI tokens to signal productivity. In an internal email obtained by 404 Media, Executive Vice President Jay Parikh said the company is not optimizing for tokenmaxxing and wants staff to focus on meaningful outcomes for customers and the business.
Parikh said Microsoft will update internal guidelines and manage token spending with the same discipline as other important resources. Starting July 2026, departments will have AI token budget targets, and employees can track their own AI usage costs. Microsoft also plans to switch its internal AI models to the lower-cost OpenAI GPT-5.6.
The move follows similar problems at other tech firms. Uber reportedly exhausted its annual AI coding token budget in four months and imposed per-tool monthly caps of 240,000 yen per employee.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.
Sources
- GIGAZINE Microsoftが従業員によるAI利用を制限する動きに出る