Overview
- Jay Parikh, CoreAI’s executive vice president, sent an internal memo this week telling engineers that “tokenmaxxing is not what we are optimizing for” and linking to updated Copilot rules.
- The company set OpenAI’s GPT‑5.6 Sol as the default model for internal GitHub Copilot use to get better value from token spending.
- Microsoft has required every division to carry an AI token budget target and added an internal dashboard that shows individual employee token spending.
- CoreAI has not imposed hard team or per‑employee caps yet and told staff to discuss large, token‑heavy projects with managers before proceeding.
- The change follows a broader industry trend to curb runaway inference bills after agentic tools drove token use up, and it will reshape model routing, vendor choices, and how engineers balance cost with product impact.