AI-use disclosure for Anthropic apps

Name Anthropic

Anthropic is in the ai api category. On sites we scan, it shows up through api.anthropic.com. Its role is to generate text from a Claude model.

Data that may be processed: Prompts and other content your app sends to the Anthropic API. The vendor describes the location as: United States (Anthropic). Their public privacy or data-processing terms are at https://www.anthropic.com/legal/privacy.

Say whether people submit prompts or files to Anthropic, and do not claim the provider has disabled training unless you have checked that setting yourself. Model output is not professional advice.

Anthropic's API host is api.anthropic.com. If a Claude call runs only on your server, a browser scan will not see it. If the key or the request is in client code, the disclosure has to say that prompts leave the browser for Anthropic.

A disclosure is not a model card

You do not have to publish your prompt. You do have to say which provider receives user content, whether people are asked to submit that content, and that output can be wrong. TermsPack's AI-use page is one of the five documents in the pack.