Microsoft executive Jay Parikh instructed developers this week to default to OpenAI's GPT-5.6 Sol model when working on AI coding projects. The directive, contained in a memo to employees viewed by CNBC, aims to maximize efficiency from the company's token investment. Parikh, executive vice president of Microsoft's CoreAI engineering group, wrote that shifting more workloads to OpenAI models helps the company get greater value from its token spending, leveraging intellectual property rights from Microsoft's early investment in OpenAI.
Microsoft Cites Token Investment Efficiency as Rationale
Parikh told staffers to default to OpenAI's flagship GPT-5.6 Sol when working in the GitHub Copilot coding tool and use that model most of the time. "Internally, shifting more workloads to OpenAI models helps us get greater value from our token investment," Parikh wrote in the memo. Tokens measure the scale of AI processing, with one token equal to about three-quarters of a word. While Microsoft has built its own artificial intelligence programming model and gives cloud customers access to over 11,000 models, including from Anthropic, the company wants staffers to take advantage of valuable intellectual property rights that come from the software giant's early investment in OpenAI. OpenAI released GPT-5.6 Sol in July.
A Microsoft spokesperson said by email that the company periodically updates "the default model settings in our internal tools to balance performance and efficient use of resources." The spokesperson wrote that Microsoft has set OpenAI's GPT-5.6 Sol as the default for Microsoft's internal use of GitHub Copilot while continuing to offer a range of model options that can be selected at any time by engineers. The CoreAI group will adjust defaults as models and products change, Parikh wrote.
Microsoft Extended OpenAI IP Rights Through 2032
The guidance comes nine months after OpenAI completed a corporate restructuring that included extending IP rights for Microsoft through 2032. Microsoft said in April it was ceasing revenue-sharing payments to OpenAI. Parikh's group includes GitHub, Visual Studio and Visual Studio Code.
Hyperscalers Face Pressure on AI Capital Expenditures
Efficiency in AI spending is becoming increasingly important across corporate America after a brief era of so-called tokenmaxxing, when developers were encouraged to run up large token bills without worrying about their output. Capital expenditures from Microsoft, Amazon, Alphabet and Meta are expected to top $700 billion collectively this year. Across the group, free cash flow dwindled in the latest quarter and even went negative for Amazon and Alphabet. Microsoft's cash generation fell by 23% from a year earlier, a mild decrease compared to its peers.
Microsoft's stock rallied 22% last week on earnings, its strongest weekly performance since 1999. The stock is now up about 1% for the year. Microsoft CEO Satya Nadella has said companies can lower costs by separating AI services from models. Last week Microsoft said GitHub Copilot has 50 million users.
Microsoft Deepens Anthropic Partnership Alongside OpenAI Strategy
Microsoft has been tightening its relationship with Anthropic, agreeing to invest up to $5 billion in the company late last year and releasing the Copilot Cowork product containing its models. Anthropic committed to spend $30 billion on Microsoft Azure cloud services. But the companies haven't made public assurances on intellectual property. Anthropic's Claude Code software development agent only provides access to Anthropic models, while GitHub Copilot offers models from Anthropic, Google, Moonshot AI, OpenAI, xAI and Microsoft itself.
Parikh noted that while Microsoft divisions are managing AI token budgets, CoreAI has not yet established budgets for individual teams or employees. He encouraged each employee to find an example of AI spending that helped a customer or business outcome as well as one that did not. "If you have a big idea or big project that will need significant token usage, have a quick chat with your manager," Parikh wrote.
FAQ
What did Microsoft instruct developers to do this week?
Microsoft executive Jay Parikh instructed developers this week to default to OpenAI's GPT-5.6 Sol model when working in the GitHub Copilot coding tool, according to a memo to employees viewed by CNBC. The directive aims to maximize efficiency from the company's token investment.
How long do Microsoft's IP rights with OpenAI extend?
Microsoft's intellectual property rights with OpenAI extend through 2032, following a corporate restructuring completed nine months before the memo was issued. Microsoft said in April it was ceasing revenue-sharing payments to OpenAI.
How much are hyperscalers expected to spend on AI capital expenditures this year?
Capital expenditures from Microsoft, Amazon, Alphabet and Meta are expected to top $700 billion collectively this year, according to the source article.
