Source : INDIA TODAY NEWS
Microsoft has reportedly shifted to GPT-5.6 Sol – the most powerful AI model currently available from OpenAI – as the default AI model for its employees. But at the same time, the company has asked its staff to stop tokenmaxxing, and focus more on the output of their work instead.
As per a report from 404 Media, in an internal memo, Microsoft executive vice-president Jay Parikh advised employees to rethink how they used tokens with the company’s GitHub Copilot tool. “As we accelerate our use of GitHub Copilot to deliver on our goals, we all need to be aware of how we consume tokens,” Jay Parikh said. “Tokenmaxxing is not what we are optimising for. I want all of us focused on maximising outcomes that move the needle for our customers and our business.”
advertisement
In case you are unaware, tokenmaxxing refers to using the maximum number of AI tokens. Tokens act as a basic unit of measurement for AI. The more work you do, the more tokens you consume.
Microsoft is not the only company that is rethinking how its employees use AI. In recent months, companies like Uber and Amazon have also cracked down on tokenmaxxing in a bid to curb rising AI costs.
GPT-5.6 Sol is cheaper to use
The memo adds that Microsoft switched to GPT-5.6 in an effort to “get greater value from our token investment.” As per a report from Sources, GPT-5.6 Sol is cheaper to use compared to Anthropic’s Claude models.
Jay’s email also links to updated internal Copilot guidelines stating that, as of July 2026, Microsoft divisions will have an “AI token budget target,” and that employees can track their individual AI spending. Though the Microsoft executive added that there was no target set just yet.
The guidelines shared by Jay Parikh state, “The data shows that many engineers spend in the range of hundreds of dollars a month to a few thousand dollars in tokens.”
Parikh told employees the company was “not optimising for fewer tokens” but for “more impact per token,” and said the changes were not meant to slow Microsoft’s push to become “AI-first.”
Keep in mind that Microsoft has invested billions of dollars into AI, including building its own MAI family of models. As per reports, Microsoft may also be trying to reduce AI costs by routing Microsoft 365 AI prompts to MAI models instead of those from Anthropic and OpenAI. Currently, Microsoft offers over 70 AI tools under the Copilot name for enterprise and consumers.
Companies are rethinking AI costs?
Microsoft has not introduced the policy because it is short of money, according to the report. Its latest earnings showed revenue, operating income and net income all rose and beat Wall Street expectations. Even so, the company is now joining other large firms that have started to curb expensive AI usage.
As we mentioned earlier, companies such as Amazon, Adobe, Atlassian and Citi have also moved to throttle employee AI use, while Meta imposed token budgets and later shut down ‘Claudeonomics,’ an internal leaderboard that had employees competing to burn tokens.
The wider shift has also been visible elsewhere in the industry. At Google I/O in May, Sundar Pichai said many companies had already exhausted their annual token budgets and presented Gemini 3.5 Flash as a cheaper alternative.
OpenAI CEO Sam Altman has also chimed into this discussion. During an enterprise event in June, Sam revealed that companies were now more concerned about AI costs. He said, “It’s kind of a meme now, ‘My company spent my entire 2026 budget in Q1, can you make this more efficient?’”
– Ends
SOURCE :- TIMES OF INDIA
