AWS CEO Defends Backing OpenAI and Anthropic
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Matt Garman told the HumanX conference that AWS's simultaneous backing of OpenAI ($50B) and Anthropic ($8B) reflects a long-standing practice of competing with partners while promising them no unfair advantage.
- AWS has managed partner-competition since its earliest years — even rival Oracle sells its database services on AWS, a dynamic Garman called 'radical' when AWS launched in 2006 but now standard.
- Anthropic's $30 billion funding round in February included at least a dozen investors also backing OpenAI, including Microsoft, OpenAI's main cloud partner — showing overlapping AI investment is industry-wide, not unique to Amazon.
- AWS's OpenAI bet was nearly existential because both OpenAI and Anthropic models were already available on Microsoft Azure, AWS's biggest rival.
- Cloud giants are now offering AI model-routing services that let customers automatically pick different models for different tasks — a mechanism that, as Garman acknowledged, lets Amazon and Microsoft slip their homegrown models into usage.
- Garman argued that routing different models to different tasks — one for planning, another for reasoning, a cheaper one for code completion — is 'where the world will go.'
Why it matters: The cloud wars have become existential in the AI era: with both OpenAI and Anthropic's models already on Microsoft Azure, AWS faced the prospect of being cut out of the most important AI distribution channel of the decade. Garman's defense — that competing with partners while backing them is old hat — masks a deeper power play, since model-routing services give Amazon and Microsoft a structural lever to push their own homegrown models into customer workflows regardless of which lab built the frontier model.
Ask SkimNews
