Your content and model output
You retain all rights, title, and interest in the prompts, source code, files, configurations, and other content you submit (“Inputs”). You grant us a limited licence to process Inputs solely to operate and secure the Services for you — including indexing them for retrieval, generating repository dependency graphs, and transmitting them to designated inference endpoints to produce responses.
As between you and us, you own the responses the Services generate from your Inputs (“Outputs”), to the extent such rights can be assigned under applicable law. We assign to you whatever rights we hold in Outputs generated for you.
You are responsible for evaluating Outputs before relying on them. AI systems are probabilistic: Outputs may be inaccurate, incomplete, outdated, or unsuitable for your purpose, and similar Inputs from other customers may produce similar Outputs. Do not rely on Outputs as professional advice, and review generated code before running it in production.
We do not use your Inputs or Outputs to train foundation models by default. See “Model training and data sharing” below for our opt-in framework and safeguards.