Microsoft Weighs Legal Action Over Amazon’s $50 Billion OpenAI Partnership

TradingKey - Microsoft is assessing legal action regarding a cloud partnership deal between Amazon and OpenAI valued at up to $50 billion, according to the Financial Times. The core of the dispute involves "Frontier," an enterprise-grade AI agent platform co-developed by the two parties, whose operating model on AWS is being questioned by Microsoft for potentially violating existing contractual terms.

[Source: Financial Times]

According to previous arrangements, while Microsoft relinquished its status as the exclusive cloud provider when it approved OpenAI's restructuring last year, it retained a key provision: all requests to call OpenAI models via API must be processed through the Azure ecosystem.

However, by building a "Stateful Runtime Environment" (SRE) on Amazon Bedrock, Amazon and OpenAI have enabled AI agents to directly call data stored on AWS and execute cross-tool tasks, which is viewed as a technical path to circumvent that restriction.

Sources familiar with the matter revealed that Microsoft believes this solution, even if technically sound, may violate the core spirit of the agreement and does not rule out legal action to protect its interests. In contrast, OpenAI contends that Frontier does not provide services in the form of a traditional API and therefore does not cross the agreement's red lines.

Notably, internal Amazon documents show the company has strictly limited related terminology to prevent external parties from interpreting the service as "providing direct access to OpenAI models," thereby mitigating potential legal risks. Currently, the three parties are still in negotiations, attempting to reach an agreement before the official launch of Frontier.

Analysts noted that should the dispute escalate into a legal battle, it could not only reshape the collaborative boundaries of cloud computing and AI model distribution but also introduce uncertainty into OpenAI's potential IPO process.