Market moves from cloud exclusivity to a multicloud AI strategy and a substantially accelerated competitive environment.
Key Takeaways
- OpenAI is shifting from cloud exclusivity to “platform optionality,” but Microsoft remains the primary partner. The key change isn’t a breakup—it’s a revised structure that permits OpenAI to deploy and sell on multiple clouds (including AWS) while preserving Microsoft’s preferential rights, licensing, and revenue-sharing through 2030.
- AWS is using Bedrock to neutralize Azure’s former OpenAI advantage in enterprise procurement. By making OpenAI models and managed, OpenAI-powered agent capabilities available inside AWS’s governed environment (security/compliance/billing controls), AWS can credibly pitch “OpenAI on AWS” with enterprise-grade guardrails.
- Multi-cloud deployment can widen customer choice and reduce lock-in. However, regulators continue to scrutinize cloud–AI partnerships in concentrated, capital-intensive infrastructure markets for potential foreclosure risks, higher switching costs, and control of critical inputs like compute.
OpenAI’s expanded partnership with Amazon Web Services marks a significant shift in the relationship between major AI developers and hyperscale cloud providers. With OpenAI’s April 2026 decision to amend its exclusive cloud arrangement with Microsoft, OpenAI’s most advanced models are now available to enterprise customers through AWS’s Bedrock platform, alongside new OpenAI-powered agent services and jointly developed infrastructure tools.
According to announcements from Amazon and OpenAI, the partnership brings OpenAI’s frontier models, Codex coding tools, and managed AI agents into AWS’s enterprise AI ecosystem.
The partnership comes with substantial financial and infrastructure commitments. Amazon has agreed to invest up to $50 billion in OpenAI and to supply two gigawatts of compute capacity, using AWS’s Trainium accelerators. Separately, OpenAI has committed tens of billions of dollars in cloud spend to AWS for training and inference workloads.
From AI Exclusivity to Platform Options
For several years, Microsoft Azure was the exclusive commercial cloud provider for OpenAI’s models, a relationship both companies credited with accelerating the commercial adoption of generative AI. That exclusivity is now over.
According to OpenAI and Microsoft, amendments announced in late April 2026 permit OpenAI to deploy and sell its models on multiple cloud platforms—including AWS—while preserving Microsoft’s status as OpenAI’s primary partner. Microsoft retains long-term licenses to OpenAI intellectual property, preferential access to new models where technically feasible, and a capped revenue-sharing arrangement through 2030. OpenAI products will continue to ship first on Azure unless Azure cannot support necessary capabilities.
This restructuring reflects the growing scale of AI infrastructure requirements rather than a breakdown in the Microsoft–OpenAI relationship. According to statements by OpenAI executives, enterprise customers increasingly require multi-cloud availability, regional flexibility, and redundancy—demands that are difficult to satisfy under exclusive cloud arrangements.
AWS’s Competitive AI Positioning
AWS has moved quickly to fill the gap left by the end of OpenAI’s Azure exclusivity. According to Amazon, OpenAI models are now available through Amazon Bedrock, AWS’s managed platform for deploying foundation models under unified security, governance, and billing controls. AWS has also launched Bedrock Managed Agents, allowing enterprises to deploy OpenAI‑powered agents entirely within AWS infrastructure.
According to AWS, inference and agent operations run inside the customer’s AWS environment, addressing common enterprise concerns about data control, auditability, and regulatory compliance. From a competition perspective, this places AWS on substantially equal footing with Azure for customers seeking OpenAI-based AI services.
Cloud Market Structure: Concentration Without Majority Control
Despite its scale, AWS does not control a majority of the global cloud infrastructure market. According to Synergy Research Group and other industry analysts, AWS accounts for approximately 30 percent of worldwide cloud infrastructure spending, compared with roughly 20–22 percent for Microsoft Azure and about 13 percent for Google Cloud. Collectively, these three providers account for roughly two thirds of global cloud infrastructure revenue, but no single provider exceeds a majority share.
This market structure helps explain why OpenAI’s shift to multi-cloud deployment is competitively significant. Prior to April 2026, access to OpenAI’s models served as a differentiating feature for Azure in enterprise procurement decisions. According to customer and analyst commentary, that differentiator diminished once OpenAI models became available through AWS under comparable governance and security conditions.
What Are U.S. Antitrust Authorities Doing About Deals in the AI Market?
To date, no U.S. antitrust authority has filed a complaint challenging the AWS–OpenAI partnership itself. However, the arrangement sits squarely within an area of active enforcement scrutiny.
According to the Federal Trade Commission, partnerships between dominant cloud service providers and leading AI developers raise potential competition concerns even where they do not involve a traditional acquisition. In a January 2025 staff report issued following a compulsory Section 6(b) study, the FTC examined relationships involving Microsoft and OpenAI, Amazon and Anthropic, and Google and Anthropic.
According to the FTC’s report, such partnerships may increase switching costs, affect access to critical inputs such as compute resources, and grant cloud providers access to competitively sensitive information unavailable to rivals. The Commission emphasized that these findings were intended to inform future enforcement decisions, signaling continued scrutiny rather than immediate litigation.
The Department of Justice’s Antitrust Division has likewise indicated interest in AI infrastructure markets. According to public remarks and reporting during 2024–2025, DOJ officials have been examining whether vertical integration across compute, capital, and AI model deployment could enable foreclosure or raise rivals’ costs, though no enforcement action has been announced specific to AWS and OpenAI.
How Are European Agencies Asserting Their Authority in the AI Deal Arena?
Agency scrutiny is more advanced in Europe. According to the European Commission, formal market investigations under the Digital Markets Act were opened in November 2025 to assess whether Amazon Web Services and Microsoft Azure should be designated as “gatekeepers” for cloud computing services.
According to the Commission, these investigations focus on whether cloud providers function as chokepoints for business users and whether competition is restricted by interoperability constraints, data access limitations, tying, or bundling. These issues have become increasingly important as AI development increasingly depends on clouds.
Separately, the Commission has launched broader antitrust probes into AI partnerships between large technology companies and AI developers, including arrangements involving OpenAI. These probes examine whether preferential or exclusive agreements could foreclose competition or disadvantage independent and European AI developers.
What is the Competitive Significance of the AWS-OpenAI Deal?
From a competition perspective, OpenAI’s move toward multicloud availability cuts against concerns about exclusive vertical alignment between dominant platforms and upstream AI developers. By reducing dependence on a single hyperscaler, OpenAI gains greater bargaining leverage, while enterprise customers gain additional choice.
At the same time, the scale of the underlying commitments—tens of billions of dollars in investments, long-term compute contracts, and preferential access arrangements—underscores the highly concentrated and capital-intensive nature of the AI infrastructure layer. According to regulators on both sides of the Atlantic, this concentration continues to warrant close monitoring even as formal exclusivity diminishes.
Looking Ahead: AI Competition
The AWS–OpenAI partnership reflects a broader normalization within the AI sector. As frontier AI providers mature and demand accelerates, exclusive cloud relationships increasingly give way to multicloud strategies driven by scale, resilience, and regulatory risk.
Whether this shift ultimately strengthens competition or merely redistributes bargaining power among a small number of incumbents, will depend on how durable and open these multicloud arrangements remain as AI markets continue to evolve.
Frequently Asked Questions
No. As of May 2026, no U.S. antitrust authority has filed a lawsuit challenging the AWS–OpenAI partnership, though regulators are actively studying similar AI–cloud relationships.
No. According to OpenAI and Microsoft, the companies remain long-term partners, but OpenAI is no longer restricted to a single cloud provider.
According to regulators and industry analysts, multi-cloud deployment can reduce lock-in, lower switching costs, and give customers greater bargaining power.
Yes. According to the European Commission, AWS and Azure are subject to ongoing Digital Markets Act investigations, and AI partnerships are part of broader competition reviews.
Possibly. According to the FTC and EU competition authorities, current studies and market investigations are intended to inform potential future enforcement if competitive harms are identified.
A hyperscaler is a cloud service provider that operates data-center infrastructure at massive scale, typically offering on-demand computing, storage, networking, and platform services across multiple global regions. The largest hyperscalers—Amazon Web Services, Microsoft Azure, and Google Cloud—serve millions of business customers and account for a majority of global cloud infrastructure spending. They are significant in AI markets because they control several critical inputs simultaneously, including compute capacity, specialized AI chips, global distribution networks, and enterprise customer relationships. According to competition authorities, this concentration can create high barriers to entry and give hyperscalers substantial influence over how AI models are trained, deployed, and commercialized. As a result, partnerships between hyperscalers and leading AI developers have become a focal point for antitrust and regulatory scrutiny.
Yes. Mogin Law LLP is among organizations tracking mergers, acquisitions, alliances, and partnerships among tech companies in the races for position in the artificial intelligence space, or in their existing spaces using the power of AI. See the interactive Mogin Law AI Deal Table.
Mogin Law LLP is a highly regarded antitrust boutique with a considerable track record of success in representing or advising companies in highly competitive spaces and with regard to mergers and acquisitions in their industry. Write to Mogin Law to set up a consultation.
Sources
Public statements and reporting from OpenAI, Amazon Web Services, Microsoft; reporting by CNBC, VentureBeat, and CRN; FTC AI Partnerships Staff Report (Jan. 2025); European Commission DMA market investigation announcements (Nov. 2025).