Microsoft is reportedly looking to reduce its reliance on OpenAI for the AI models powering its Microsoft 365 Copilot assistant. Citing cost and speed concerns, Reuters reported yesterday that Microsoft is developing in-house, smaller AI models to make its flagship AI assistant more efficient. The company is also said to be looking at third-party AI models to diversify the technology behind Microsoft 365 Copilot.
“In addition to training its own smaller models including the latest Phi-4, Microsoft is also working to customize other open-weight models to make 365 Copilot faster and more efficient,” Reuters said citing sources familiar with the efforts. “The goal is to make it less expensive for Microsoft to run 365 Copilot, and potentially pass along those savings to the end customer, one of the sources said.”
Sign up for our new free newsletter to get three time-saving tips each Friday — and get free copies of Paul Thurrott's Windows 11 and Windows 10 Field Guides (normally $9.99) as a special welcome gift!
"*" indicates required fields
In recent years, Microsoft invested billions of dollars into OpenAI to get an exclusive license to use and customize OpenAI’s GPT-4 model, which powers its Microsoft 365 Copilot offering for commercial customers. The AI assistant is priced at $30 per user per month on top of a qualifying Microsoft 365 plan, and it’s integrated into various Microsoft 365 apps. Organizations can also add agents to Copilot to automate business processes.
In a statement shared with Reuters, a Microsoft spokesperson said “We incorporate various models from OpenAI and Microsoft depending on the product and experience.” The company also said that OpenAI remains Microsoft’s main partner to work on the most advanced AI models coming to the market.
While Microsoft’s partnership with OpenAI allowed the company to get its hands on cutting-edge AI models, the launch of a new “Microsoft AI” subsidiary earlier this year probably marked a turning point in an increasingly complex relationship. Led by DeepMind and Inflection co-founder Mustafa Suleyman, this new AI division is now developing proprietary AI models. The new version of Copilot for consumers that Microsoft launched in October actually uses a combination of in-house AI models and OpenAI models.
According to Reuters, Microsoft CEO Satya Nadella and other Microsoft leaders are keeping a close eye on the company’s efforts to reduce its reliance on OpenAI. In addition to the consumer version of Microsoft Copilot no longer relying as much on OpenAI technology, Microsoft also announced in October that GitHub Copilot was getting support for new AI models from Anthropic and Google. At the time, the company acknowledged that developers should be able to “decide which models work best for them.”