Amazon has placed billions of dollars into both Anthropic and OpenAI - two companies competing directly for the same enterprise AI contracts. AWS CEO Andy Jassy says this isn't a problem, and his reasoning is straightforward: Amazon has been doing this kind of thing for decades.
The logic draws on AWS's core business model. Netflix runs on AWS infrastructure while competing with Amazon Prime Video. Walmart uses AWS while competing with Amazon retail. Amazon has spent years building internal walls between competitive intelligence and its commercial relationships, and Jassy argues that same discipline applies to its AI investments.
That parallel has limits. Anthropic's deal with AWS includes a commitment making AWS its primary cloud and AI training provider. This isn't a passive equity stake - it's a structural dependency baked into both companies' operations. The same appears true of OpenAI's AWS relationship.
The real tension: AWS runs Bedrock, its platform for enterprises to plug different AI models into their applications. When Bedrock decides which models to feature, price favorably, or prioritize in recommendations, those decisions happen inside the same company that owns equity in the models being evaluated. Amazon says these processes are kept separate. That may be true. But "trust us, we have walls" is a harder sell when the financial entanglements are this deep.
For individual users, none of this changes how Claude or ChatGPT works day-to-day. For enterprise buyers making long-term infrastructure decisions on AWS, it's worth asking whether the platform's model recommendations come with a structural bias baked in.