OpenAI has sought to dispel growing speculation that its relationship with Microsoft is weakening, announcing that its newly launched GPT-5.6 model will serve as the “preferred model” for Microsoft 365 Copilot, the AI assistant embedded across Microsoft’s productivity software.
The announcement came during OpenAI’s launch event for GPT-5.6 on Thursday and follows recent reports suggesting Microsoft has been relying more heavily on its own artificial intelligence models to reduce costs, prompting questions about the future of one of the technology industry’s most influential partnerships.
In a blog post accompanying the launch, OpenAI said GPT-5.6 will power AI capabilities across Microsoft’s suite of productivity applications, including Word, Excel, PowerPoint and Cowork, reinforcing the companies’ continued collaboration despite mounting speculation about diverging AI strategies.
“Our partnership with Microsoft has always been about bringing the benefits of advanced AI to more individuals and organizations, and we’re excited to continue building on that shared commitment,” OpenAI said.
The announcement appears intended to reassure customers and investors that OpenAI remains central to Microsoft’s flagship AI offerings.
However, the practical implications of the designation remain unclear.
OpenAI did not specify what the ” preferred model being” means from a technical or commercial standpoint, nor did it indicate whether Microsoft will continue deploying its own proprietary AI models alongside GPT-5.6.
The development follows a Bloomberg report earlier this week that said Microsoft has increasingly begun replacing some OpenAI technology with internally developed artificial intelligence models known as MAI.
According to the report, Microsoft’s MAI models are already being used to power features in applications such as Word and Excel as the company seeks to reduce the substantial costs associated with operating large language models.
The report fueled renewed speculation that Microsoft and OpenAI, whose close alliance has reshaped the artificial intelligence industry over the past several years, may gradually be pursuing more independent strategies.
Since Microsoft’s multibillion-dollar investment in OpenAI, the two companies have worked closely to integrate OpenAI’s technology across Microsoft’s products, making services such as Microsoft 365 Copilot among the most prominent commercial deployments of OpenAI’s models.
More recently, however, Microsoft has significantly expanded its own AI research and development efforts. The company has introduced its MAI family of models while also developing specialized AI systems designed to reduce dependence on third-party providers and lower computing costs.
At the same time, OpenAI has increasingly broadened its commercial relationships beyond Microsoft, partnering with additional cloud providers, enterprise customers and hardware companies as it scales its own business. Those developments have led industry observers to question whether the partnership is evolving from an exclusive strategic alliance into a more conventional commercial relationship between two increasingly independent AI companies.
OpenAI’s latest announcement suggests that, at least for now, its models will continue to play a prominent role within Microsoft’s productivity ecosystem.
But the company’s statement does not directly contradict reports that Microsoft is incorporating more of its own AI technology into products where it believes doing so offers technical or economic advantages. Rather than indicating an either-or choice between OpenAI and Microsoft’s in-house models, the latest developments suggest Microsoft may increasingly adopt a hybrid approach, using OpenAI’s frontier models where they deliver the strongest performance while deploying internally developed models for specific applications or cost-sensitive workloads.
Such a strategy would allow Microsoft to balance access to OpenAI’s latest advances with greater control over operating expenses and product development.
Maintaining a leading role in Microsoft 365 Copilot remains strategically significant for OpenAI. Microsoft’s productivity applications serve hundreds of millions of users worldwide, making the software suite one of the largest enterprise distribution channels for generative AI.
Keeping GPT-5.6 at the center of those products helps preserve OpenAI’s visibility and bolsters its position in the highly competitive market for enterprise AI services.
While the “preferred model” designation signals that the partnership remains active, it leaves unanswered broader questions about how responsibilities and technologies will be divided as both companies continue investing heavily in their own AI capabilities. The announcement therefore appears less like a reversal of reports about Microsoft’s growing use of proprietary models and more like confirmation that the two companies continue to collaborate even as each pursues increasingly independent AI strategies.
