Privacy notice

Last updated June 12, 2026.

XData indexes public expert content. The transcript corpus is private during the preview; browser and agent access require OAuth or an issued API key.

Information we process

• OAuth identity: email, name, provider subject.
• Access requests: email, name, organization, intended use.
• Authenticated access: API key records, usage counters, rate-limit state.
• MCP requests, including searches and public ingestion URLs.
• Your theme preference is stored only in your browser.
• Public text and metadata: podcasts, videos, interviews, feeds, newsletters, webpages.

How content is handled

Audio is downloaded, transcribed on our hardware, then deleted before the content record is stored. We retain text, source URLs, metadata, retrieval chunks, embeddings, extracted statements, and an audit record of every removed advertising span.

Processors

Providers include Render, WorkOS, and OpenRouter. Transcription uses our hardware, not third-party speech APIs. Queries are not used to train models.

Retention and access

Account and usage records remain while access is active or needed for security and abuse prevention. For access, correction, deletion, or source removal, see the support page.