Microsoft has ‘good news’ and ‘bad news’ for engineers using AI coding tools

Microsoft has told engineers to curb AI token spending, with EVP Jay Parikh writing in an internal memo that “tokenmaxxing is not what we are optimizing for.” The company has made OpenAI’s cheaper GPT-5.6 the default internal model, set AI token budget targets for every division from July 2026, and given employees dashboards to track individual spend on GitHub Copilot. Some engineers were burning thousands of dollars monthly.