The balance of power in the global AI industry is subtly shifting. Microsoft and OpenAI have officially announced significant revisions to their partnership agreement today, aiming to provide greater flexibility for both companies in a rapidly changing market.

The licensing model shifts from exclusive to non-exclusive

According to the latest revised terms, Microsoft retains the intellectual property license for OpenAI's models and products, which will remain valid until 2032. However, this license has transitioned from an "exclusive" model to a "non-exclusive" one. This means that OpenAI now has full autonomy to offer its entire product line to customers through any cloud service provider.

Although the licensing model has changed, Microsoft continues to maintain its position as OpenAI's primary cloud service partner. Unless Microsoft's Azure cloud platform cannot meet specific functional requirements, OpenAI's new products will continue to be launched primarily on Azure. Collaboration between the two companies in data centers and chip development will also continue.

Financial arrangements clarified and roles transformed

In financial terms, this agreement has reached several "liberalization" consensuses. Microsoft will stop paying revenue shares to OpenAI, while OpenAI's revenue share to Microsoft will continue until 2030, with a total cap, no longer directly affected by technological developments.

Although financial interactions have been streamlined, Microsoft will still directly participate in OpenAI's long-term development as a major shareholder. Both parties stated that this adjustment has clarified operational details, retaining Microsoft's investment rights while also giving OpenAI more independence to expand in the global artificial intelligence market.