Unbundling from Microsoft, Embracing Amazon—Why Is OpenAI Doing This?

By restructuring its cooperation agreement with Microsoft, OpenAI has formally ended its long-standing exclusive cloud services binding and is embracing other cloud platforms such as Amazon, intending to break infrastructure limitations and fully compete for the enterprise-level market.

As Wall Street Insight previously reported, on Monday, Microsoft and OpenAI announced revisions to their cooperation agreement. Core terms include Microsoft ceasing revenue-sharing payments to OpenAI and changing its intellectual property license for OpenAI’s models from exclusive to non-exclusive. This historic unbundling directly triggered market volatility: Microsoft’s stock plummeted nearly 4% in pre-market trading before recovering some losses, while Amazon’s stock rose approximately 1% amid positive expectations for its multi-cloud strategy.

On the same day, Amazon announced that it had reached an agreement to bring OpenAI onto its Amazon Web Services (AWS) cloud platform, which will launch new services specifically designed to run AI agents.

For OpenAI, embracing platforms like Amazon is a key step in catching up with competitor Anthropic and breaking through growth bottlenecks in its enterprise business. Previously constrained by exclusive terms with Microsoft Azure, OpenAI was unable to serve certain enterprise customers unwilling to migrate cloud platforms. Ending this binding allows OpenAI to align with the industry trend toward multi-cloud deployments and reactivate broad potential demand.

Although OpenAI has gained crucial operational autonomy and is eager to expand its revenue footprint through Amazon Web Services (AWS), its expansion path in the multi-cloud era still faces severe commercial tests, as enterprise customers have shifted to other highly competitive AI models in recent years.

Ending the Exclusive Agreement, Reshaping the Cloud Services Value Chain

This agreement revision marks a significant restructuring of the financial relationship between Microsoft and OpenAI.

Under the new terms, Microsoft will no longer pay revenue shares to OpenAI, eliminating a fixed cost expense for Microsoft. However, as a trade-off, Microsoft has lost its moat advantage in the AI infrastructure sector, with its exclusive IP access rights broken. Meanwhile, OpenAI’s reverse revenue-sharing payments to Microsoft will continue until 2030, maintaining the original ratio but subject to a total cap.

In terms of equity, Microsoft acquired a 27% stake in OpenAI after the latter completed its for-profit restructuring last year and will remain a major shareholder participating in its growth. Reportedly, both parties engaged in months of tense negotiations last year regarding ambiguous AGI clauses in the agreement. The new agreement removes the mechanism stating that “Microsoft’s exclusive access rights terminate once the AGI threshold is reached,” eliminating a major uncertainty in their relationship. The market’s interpretation of this revision reflects a cautious assessment of whether Microsoft can maintain its core position, while Amazon is generally expected to benefit directly from this shift toward non-exclusivity.

Embracing Amazon, Intensifying the Battle for the Enterprise Market

OpenAI’s expansion into AWS is essentially an inevitable choice driven by growing computing power demands and fierce competition in the enterprise market.

Amazon previously announced an initial $15 billion investment in OpenAI, planning to introduce OpenAI technology featuring a “stateful” runtime environment on AWS’s Bedrock service. This environment retains interaction details to support more continuous usage. AWS will also provide OpenAI Frontier to help enterprises deploy AI assistants.

OpenAI executives stated in a memo that market demand for AWS’s new products was “stunning.” An OpenAI spokesperson also noted that enterprise customers seek solutions that drive actual return on investment (ROI), and the new products will enable enterprises to natively deploy agent workflows within Amazon Bedrock. To further promote this collaboration, AWS CEO Matt Garman and OpenAI executives plan to host an event for AWS customers on April 28, focusing on agentic AI technologies.

OpenAI is currently rushing to catch up with competitor Anthropic in the enterprise market. Disclosures indicate that Anthropic’s recent annualized revenue has reached $30 billion, slightly ahead of OpenAI. Consequently, OpenAI leadership has recently requested employees to reduce side projects and focus on winning more enterprise business.

Lukewarm Customer Response, Alternative Models Have Built Barriers

Despite Amazon’s active promotion of OpenAI’s inclusion, some AWS customers have reacted indifferently.

In the three years since OpenAI sparked the AI boom, many enterprises have become deeply reliant on other models. Six enterprises or consultants cooperating with AWS stated that they are satisfied with cost-effective options such as Anthropic and Amazon’s proprietary model Nova, used via Bedrock.

Adam Sandman, CEO of testing software company Inflectra, stated, “If this news had been announced a few years ago, it might have been significant, but the current market focus is truly on Qwen and DeepSeek.” He emphasized that Claude already performs better in many practical tasks such as coding.

Inflectra previously had about 10% of its customers choosing Microsoft Azure to access OpenAI, but this proportion has now dropped to less than 1%, with the vast majority of customers returning to AWS managed services using Bedrock models.

Furthermore, data control and cost-effectiveness have become significant factors hindering customers from switching to OpenAI.

Phil Christianson, Chief Product Officer at IT software company Xurrent, pointed out that transferring data to Microsoft Cloud conflicts with customers’ requirements to keep data within AWS. As the gap between frontier models narrows, the motivation to switch platforms is insufficient. Cris Daniluk, CEO of Rhythmic Technologies, which provides AWS consulting services, also mentioned that while the new runtime environment offers technical advantages, being forced to bind to OpenAI products can sometimes be troublesome. As the advantages of multi-cloud layouts in reducing latency and optimizing performance become apparent, OpenAI must still prove its irreplaceable commercial value to regain influence in a market already penetrated by competitors.

Risk Warning and Disclaimer

The market involves risks; investment requires caution. This article does not constitute personal investment advice, nor does it consider the specific investment objectives, financial status, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investment decisions made based on this content are at the user’s own risk.