Does Perplexity Track You?
Search History and AI Privacy

Published July 11, 2026 · Updated August 18, 2026 · Facts last verified August 18, 2026

Short answer: yes — on the consumer product, Perplexity is built to track you. Your search history is stored and tied to your account, it is used to train and improve Perplexity's AI by default, and the company has been unusually open that data is central to its plans: it built a browser designed to watch what you do online and turn it into "hyper-personalized" ads. Of all the big AI search tools, Perplexity's direction leans most explicitly toward advertising.

This guide decodes what Perplexity actually collects and keeps, whether it trains on your searches, what the Comet browser changes, and how the consumer product differs from the API. Policies change, so treat this as an informed starting point and confirm the current terms for your own account.

The Short Answer: Yes on the Consumer Product — Though the Ads Story Has Changed

Perplexity is a search product, and search queries are among the most revealing data there is — what you are curious about, worried about, buying, or planning. On the consumer app, those queries are saved to your account, used to improve Perplexity's models, and handled by a company whose leadership has said plainly that it wants more of your data, not less.

That framing needs a correction that earlier versions of this article got wrong, and it runs in Perplexity's favour. In April 2025 its CEO talked openly about wanting data from outside the app to build a better profile and show ads, and that ambition shaped a lot of commentary — including ours. But Perplexity phased out the advertising it had begun testing in 2024, and as of February 2026 said it has no plans to bring ads back, on the reasoning that even labelled sponsored placements would make users second-guess the answer. So the honest position today is narrower: your searches are stored and used to improve models by default, which is a real privacy question, but "Perplexity is building an ad-profiling machine" is no longer an accurate description of where the company says it is going.

Your Search History Is Stored — and Trained On

Here is what happens to your searches on a normal consumer plan (Free, Pro, or Max):

• History is kept until you delete it. Your searches and threads stay in your account until you manually remove them.
• Account deletion has a 30-day window. Perplexity documents that account and personal data are deleted within about 30 days of closing the account, subject to narrow legal and business exceptions. Note what it does not currently document: a fixed 30-day backend window for an individually deleted conversation — an earlier version of this article asserted one, and we could not substantiate it.
• Training is on by default. Perplexity uses your data to train and improve its AI unless you opt out. The control is the "AI data retention" toggle on the Preferences page, reached from All Settings — Account and Preferences are sibling pages rather than one nested in the other, and another Perplexity help page calls the same function "AI Data Usage". Turning it off applies only to future data — it does not necessarily remove what has already been used.
• Trackers and analytics. Perplexity's privacy notice discloses cookies, pixels and related analytics technologies, along with controls for targeted advertising — the ordinary web-tracking layer sitting on top of the AI itself.

The Comet Browser: What It Actually Sends

Perplexity's Comet browser launched on 9 July 2025, and it is where the ad ambition was most visible. In April 2025 the CEO said the company wanted data from outside the app — that purchases, hotels, restaurants and general browsing reveal far more about a person than search alone — and floated using that context to build a better profile and show ads in Discover. That is a real quote and worth knowing.

It is not, however, a description of how Comet documents its behaviour today, and the difference matters if you are deciding whether to install it. Perplexity's current Comet privacy documentation says browsing history, tabs, cookies, passwords and local files stay local by default, with only the context required to serve a given request transmitted. Comet also ships with built-in ad blocking. Treat the browser as a much larger surface than a search box — an assistant sitting inside your browsing has access to more of your life by construction — but do not assume everything you do in it is uploaded, because the documentation says otherwise.

Two feature names get conflated, including in earlier versions of this article. Personal Search is a Comet feature that works over your locally available browsing history to find things you have already visited. Reaching into Gmail, calendar or files is a separate connector you grant explicitly. They deepen the same profile, but they are different decisions and you can take one without the other.

The Controls the Article Used to Skip

Perplexity gives you more levers than the training toggle, and a piece asking whether it tracks you should say so.

• Incognito threads stay out of your history and out of Memory, and expire after 24 hours — the shortest window among the major assistants.
• Memory has its own controls, including whether search history feeds it, and saved memories you can read, edit and delete. Cleared-memory logs can persist for up to about 30 days.
• Logged-out threads are kept on a shorter clock (around 14 days) than account-linked ones.
• Brain, released to Max subscribers on 27 July 2026, is the newest and largest of these: a self-improving memory that builds wiki-like context about your people, projects, preferences and prior work. It is the feature most worth a deliberate decision, because it is designed to accumulate rather than answer and forget.

Consumer Perplexity vs the API: What Applies Where

As with other providers, how your data is treated depends on how you reach Perplexity.

• Consumer app (Free/Pro/Max): search history stored, trained on by default, and part of a data-and-ads business direction.
• Enterprise: Perplexity states that enterprise data is never used to train or fine-tune its models.
• The Sonar API (for developers): Perplexity maintains a Zero Data Retention policy — prompts and responses sent through the API are not stored.

So the honest answer to "does Perplexity track you?" splits cleanly: on the consumer product, yes and increasingly so; through the API, your searches are not stored or used for training at all.

What Happens When You Delete — and How to Limit Tracking

Deleting a search removes it from your history, but a copy can persist in Perplexity's backend for around 30 days, and deleting your whole account removes your data within about 30 days. If you use the consumer product, a few steps reduce your exposure:

• Turn off the AI data retention toggle so future searches are not used for training.
• Use Incognito for anything sensitive — those threads skip history and Memory and expire in 24 hours.
• Review Memory and, on Max, Brain: read what has accumulated and delete what you did not intend to store.
• Delete searches you do not want kept, and clear your history periodically.
• Be cautious with the Comet browser and Personal Search — they widen how much of your activity is collected.
• Keep genuinely sensitive questions out of an account-linked search tool entirely.

Is Perplexity Private Enough for Sensitive Research?

Perplexity is excellent at what it does — fast, sourced answers from the live web. But its consumer privacy posture, and especially its stated move toward tracking and advertising, make it a poor place for sensitive research: medical worries, legal questions, financial situations, anything you would not want built into an ad profile. For that kind of work, the API and enterprise terms are far stricter, and keeping identifying details out of the query matters more than any setting.

How Secret Chat AI Uses Perplexity Differently

The most direct way to stop worrying about what Perplexity tracks is to stop being identifiable to it in the first place. That is what Secret Chat AI is built for: it is an anonymizer and depersonalizer. Its whole design is no profile, no association, no training — it makes you unlinkable to the providers, so you search Perplexity as a stranger. (On Secret Chat, Perplexity is even available to free users.)

• No profile of you. Registration takes an email — used only for account access and payment, never to store or associate your searches with you — with no name, phone number, or behavioral profile built across your queries.
• No search associated with you. Secret Chat reaches Perplexity through its API on your behalf, so Perplexity sees the gateway, not you — no identity, no account, and no IP address attached. The consumer tracking, ad profiling, and training-by-default do not apply.
• Nothing kept under your name. Your history lives in your browser's local storage, not a Perplexity account, and each message can produce a Session Privacy Report (PDF).

This reframes the whole tracking question: even where any data is retained at the provider, what is kept is not linked to you. The one honest boundary is that Secret Chat AI is an anonymizer, not a content filter. Perplexity still has to read your query to answer it, and Secret Chat does not encrypt or strip the words you send — it removes who you are from the request, not what the request says. So keep names and identifiers you would not want a provider to see out of your searches. You can also switch between Perplexity and other private AI models in one place.

Table: Perplexity Data at a Glance

Frequently Asked Questions

1. Does Perplexity save my search history?

   On the consumer app, yes. Your searches are kept while the account is active, until you delete them. Incognito threads are the exception — they stay out of history and expire after 24 hours. Closing your account triggers deletion of account and personal data within about 30 days. Through the Sonar API, prompts and responses are not stored.

2. Does Perplexity train on my searches?

   By default, yes on the consumer plans. You can opt out with the "AI data retention" toggle on the Preferences page in settings, but it only applies to data collected after you flip it, and Perplexity still processes your data to run the service and meet legal obligations. Enterprise data and API traffic are not used for training.

3. Is Perplexity building a browser to track me?

   Comet launched on 9 July 2025, and in April 2025 Perplexity's CEO did talk about using out-of-app data to build a better profile and show ads. Two things have changed since: Perplexity phased out its advertising tests and said in February 2026 that it has no plans to resume them, and Comet's current documentation says history, tabs, cookies, passwords and local files stay local by default, with only request-required context sent. It remains a much larger surface than a search box, but the old description of it as build detailed profiles for personalized advertising. It is optional, but it reflects the company's data-and-ads direction.

4. How do I stop Perplexity from using my data?

   Turn off the AI data retention toggle, use Incognito threads for sensitive questions, delete your search history, review what Memory and Brain have accumulated, be deliberate about Comet and any connectors you grant, and keep confidential details out of your queries. For the strongest protection, use the API or a privacy-focused interface rather than a personal account.

5. Does Secret Chat AI let Perplexity track me?

   No. Secret Chat AI reaches Perplexity anonymously through its API — no profile, account, or IP attached, and your email (used only for account access and payment) is never associated with your searches. Your history stays in your browser. Perplexity still processes each query to answer it, so remove identifiers you would not want it to see.

Conclusion

So, does Perplexity track you? On the consumer product, yes in the ways that matter — your search history is stored and used to improve models by default, ordinary web trackers are in play, and Memory and Brain are designed to accumulate context about you. But be fair about the direction: the advertising programme that made this look worse than its peers was phased out, and Perplexity said in February 2026 it has no plans to resume it. Opting out of AI Data Retention and skipping Comet help, but the real protections come from the API and enterprise terms, where your searches are not stored or used for training.

If you want Perplexity's fast, sourced answers without being profiled — an anonymous, API-based connection with no profile of you, local browser-based history, and a choice of models — Secret Chat AI is built for exactly that, without pretending to remove Perplexity's processing of your query.

Sources

• Perplexity — Privacy Notice
• Perplexity Help Center — Data Collection
• Perplexity Help Center — Enterprise & API data retention
• Search Engine Land — Perplexity stops testing advertising (February 2026)
• Malwarebytes — Perplexity's browser, tracking, and ads