Anthropic

Anthropic PBC is a public-benefit corporation and a frontier-model lab. Claude competes with ChatGPT and Google's Gemini for consumer chat and, more importantly, for enterprise API contracts.

Amazon has invested and made Amazon Web Services a primary cloud, with distribution through Amazon's own products. Google has also invested and offers Claude in its cloud marketplace. Two hyperscalers funding the same independent lab is unusual. It can be a competitive counterweight to Microsoft–OpenAI, or a set of exclusive claims on one lab's output.

The FTC's AI partnership inquiry covers this pattern: minority stakes, cloud credits, and rights to models that may be the real merger. No complaint had rewritten Anthropic's structure as of October 2026, but the inquiry is why the company belongs in a competition file.

Enterprise buyers like Claude because contracts can limit training on customer data. That privacy position is a product difference. It also creates lock-in once a company's eval suite, prompts, and access controls are built for one provider's API.

Dual investment by Amazon and Google raises questions about information flow and whether either cloud gets preferential model access the other customers do not.

Cloud credits that are large relative to revenue can be an exclusive-dealing discount: the lab trains where the investor is, and rivals' clouds do not see the workload.

If Claude is preinstalled or defaulted inside a major cloud console, the default-distribution theory from search cases applies again.

Switching models is technically possible and operationally expensive. Eval harnesses and safety policies do not copy-paste.

Most people will see Claude only if a vendor chooses it. The competitive harm, if any, is fewer independent model suppliers behind the apps they already use.

Companies that standardize on one API inherit price and policy risk. A safety filter change can break a workflow without a rival model already certified.