OpenAI's long-term collaboration with Microsoft has undergone a significant turning point today. According to the latest revised terms, both parties have officially updated their AI technology cooperation agreement, marking the third and most far-reaching reshaping of their interests and cooperation boundaries since 2025.

The core of this revision lies in breaking the previous "exclusive" restrictions. According to the new agreement, OpenAI is now allowed to provide its full range of products to all cloud service providers worldwide, no longer bound exclusively to Microsoft Azure. This change has quickly taken effect: it has been revealed that OpenAI's relevant models will officially launch on Amazon Bedrock platform in the coming weeks. This means enterprise customers will have greater flexibility and autonomy when choosing large model infrastructure.

In terms of intellectual property and licensing, the nature of the cooperation has also undergone a fundamental change. The original exclusive license has been adjusted to a non-exclusive one, with an initial term set until 2032. Although Microsoft still retains priority access to OpenAI's products to ensure its competitive advantage in end-user experience, OpenAI's intention to expand its commercial footprint is clear, as it is striving to become an independent general AI supplier, separate from any single major player.

Changes in financial terms are also notable. While the agreed-upon revenue-sharing ratio remains unchanged until 2030, this revision sets a cap on the total amount of revenue sharing. More importantly, Microsoft will no longer need to split its own AI business revenue with OpenAI in the future. This adjustment reflects Microsoft's gradual effort to strengthen the independent profitability of its AI business and reduce the payment of technical taxes to a single partner.

Industry analysts believe that these changes represent a balance point reached after the game of technological sovereignty and commercial interests between the two sides. For OpenAI, embracing other cloud service giants like Amazon helps it expand its funding sources and spread computing power risks; for Microsoft, through a non-exclusive agreement and the cancellation of revenue sharing, it also leaves more room for developing its own AI models and optimizing cost structures. Before the key time period between 2030 and 2032, the global AI market power structure may accelerate its reshuffling due to this "liberation."