Amazon Web Services chief executive Matt Garman has defended the company's $50 billion investment in OpenAI, telling the HumanX conference in San Francisco that partnering with rivals is a structural feature of the cloud business rather than a contradiction.
The remarks come after Amazon had already committed $8 billion to Anthropic, OpenAI's primary competitor, raising questions about how the company navigates backing both sides of the frontier AI market.
Garman framed the tension as a long-standing industry reality, saying AWS developed the processes to manage such conflicts during its early years, when the business recognised it could not build every cloud offering itself and would inevitably have to both partner with and compete against the same companies.
He said AWS has sometimes launched first-party products that directly rival those of its partners, but added that the company has committed to those partners that it will not give its own offerings an unfair competitive advantage.
Garman also described the OpenAI deal as carrying a defensive rationale.
Both OpenAI and Anthropic models are available through Microsoft's Azure cloud platform, and Garman said cloud providers are increasingly building model-routing services that allow enterprise customers to direct different tasks to whichever model performs best for that purpose.
"I think that is where the world will go," he said, implying that AWS needed access to OpenAI's models to remain competitive in a market where customers expect to choose from a broad menu of frontier AI systems through a single cloud relationship.
The dynamic is further complicated by overlapping investor bases across the AI sector.
Anthropic's latest funding round, which was reported at $30 billion in annualised recurring revenue disclosed this week, included multiple investors who also back OpenAI, among them Microsoft, according to TechCrunch.
Related reading
- Mistral AI raises $830 million for Paris data center
- Anthropic launches Claude Managed Agents as enterprise revenue hits $30 billion run rate
- Microsoft warns AI agents risk becoming "double agents" as it unveils security controls at RSAC
The broader picture reflects how capital and commercial relationships in frontier AI have become deeply interwoven, with the same cloud providers, technology companies and investment funds simultaneously funding competing model developers and building the infrastructure all of them depend on.
AWS has positioned itself as a neutral platform for AI model deployment, hosting models from Anthropic, and now OpenAI, alongside offerings from other developers through its Bedrock service, which lets enterprise customers access and switch between models without committing to a single provider.
The recap
- AWS chief Matt Garman defends dual AI investments.
- Amazon committed $50 billion to OpenAI following Anthropic funding.
- AWS plans model-routing services to optimise customers' AI usage.
