Tool page
Perplexity
Perplexity’s official consumer Privacy Notice, retrieved after a Cloudflare block, discloses use of data to improve or create products including AI models, and processes user prompts and queries. It explicitly does not apply to Enterprise and API, whose dedicated enterprise-privacy URL was not found.
Partial page. urllib was blocked by Cloudflare (403). A browser fetch on 2026-11-18 succeeded for the Privacy Notice (last updated October 23, 2026), Terms of Service, and Self-Serve Data Options. The Enterprise Privacy catalog URL returned 404. The consumer Privacy Notice says it does not apply to Enterprise and API offerings. Those consumer pages say Perplexity uses data to “Improve or create services and products, including our AI models,” and that they process prompts, queries, and uploads — but they do not clearly state a default training toggle or a no-training rule for ordinary queries. They do say email content is not used to train models.
Not a verdict. These are quotations from official vendor pages. They are not legal advice and not a determination that a plan is safe, compliant, or suitable for a particular kind of data.
Policy by plan
| Plan | Trains on inputs by default? | Opt-out / opt-in | Retention | ZDR | BAA / DPA | Data residency | Exceptions |
|---|---|---|---|---|---|---|---|
| Perplexity consumer services covered by the Privacy Notice (websites, Comet, mobile apps, and other services; excluding Enterprise and API) | The Notice lists among uses of data: “Improve or create services and products, including our AI models.” It also states: “We process content you submit while using our services, including prompts, queries, uploads, and any other data you choose to provide.” The page does not use a clear “trains by default / does not train” sentence for ordinary prompts. We therefore do not invent a yes/no default beyond quoting those uses. | Not stated / not verified as a model-training product control on the pages we read. The Notice and Self-Serve Data Options describe data export and other privacy choices; we did not find a clearly labeled query-training opt-out on those pages. | “We keep your personal data for only as long as necessary to fulfil the purposes in this Notice unless a longer retention period is required or permitted by law.” No numbered default period for prompts was stated. | Not stated / not verified | The Notice “does not apply to data collected through Perplexity’s Enterprise and API offerings where we act as a service provider or processor.” A DPA/BAA for those offerings was not verified because the enterprise-privacy URL returned 404. | Not stated / not verified | “We do not use the content of emails to create, train, improve, or fine-tune AI models.” That carve-out is specific to email content, not a general query-training rule. |
Source quotes
“We also use data to: Identify you when you visit our sites. Complete transactions. Improve or create services and products, including our AI models.” Perplexity consumer services covered by the Privacy Notice (websites, Comet, mobile apps, and other services; excluding Enterprise and API) · Checked 2026-11-18 · Perplexity Privacy Notice (last updated October 23, 2026)
“We process content you submit while using our services, including prompts, queries, uploads, and any other data you choose to provide.” User content processed · Checked 2026-11-18 · Perplexity Privacy Notice
“We do not use the content of emails to create, train, improve, or fine-tune AI models.” Email carve-out · Checked 2026-11-18 · Perplexity Privacy Notice
“This notice does not apply to data collected through Perplexity’s Enterprise and API offerings where we act as a service provider or processor with regard to the data collected through those systems.” Enterprise/API exclusion · Checked 2026-11-18 · Perplexity Privacy Notice
Sources checked
- Perplexity Privacy Notice
- Perplexity Privacy Policy URL (same notice)
- Perplexity Terms of Service
- Self-Serve Data Options
- Enterprise Privacy (catalog) — 404 (URL)