A Utah man logged into Perplexity AI and did what millions of people do with chatbots every day. He asked it about his taxes. He discussed his investment portfolio. He shared details about his family's finances.
He believed those conversations were private. According to a class-action lawsuit filed on Tuesday in U.S. District Court for the Northern District of California, they were not.
The complaint, Doe v. Perplexity AI Inc. (Case No. 3:26-cv-02803), alleges that Perplexity embedded "undetectable" tracking software in the code of its AI-powered search engine. The trackers, the suit claims, activate the moment a user logs into Perplexity's homepage and begin transmitting the contents of conversations to Meta Platforms and Alphabet's Google. The data allegedly flows even when a user has enabled Perplexity's "Incognito" mode.
The plaintiff is identified only as John Doe. He filed the suit on behalf of a proposed class of all Perplexity users whose data was allegedly shared without consent.
The Allegations in Detail
The complaint lays out a specific sequence. When a user visits Perplexity's homepage and logs in, tracking mechanisms are downloaded onto their device "without their knowledge or consent," according to the filing. Those mechanisms then give Meta and Google "full access to the conversations between [users] and Perplexity's AI Machine search engine."
The suit does not identify the specific tracking technology by name, but describes it as embedded within the platform's code and designed to be invisible to users. The complaint states the trackers operate in the background, forwarding chat contents in a way that enables Meta and Google "to exploit this sensitive data for their own benefit, including targeting individuals with advertising and reselling their sensitive data."
The Incognito mode claim is particularly damaging. Perplexity offers a mode that its interface suggests provides enhanced privacy for searches and conversations. The lawsuit alleges that trackers remain active in this mode, meaning the privacy promise is functionally empty.
For the named plaintiff, the stakes are personal. He told the chatbot about his household finances, tax obligations, investment portfolio, and investment strategies. If the allegations are true, that information was transmitted to two of the largest advertising companies on the planet.
Perplexity's Growing Legal Problems
This is not Perplexity's first courtroom appearance in 2026. The company, valued at approximately $20 billion after its most recent round led by Institutional Venture Partners in 2025, has faced a series of legal challenges that call into question how it handles both user data and third-party content. The company recently launched a personal computer product designed to run AI agents on a Mac Mini, expanding further into enterprise markets even as its legal exposure grows.
Amazon has sued Perplexity twice. A federal judge recently blocked Perplexity from accessing Amazon through its Comet browser, after Amazon alleged the company accessed customer accounts without permission through its automated shopping system. The company also faces a separate Amazon dispute over its "Buy with Pro" e-commerce feature, which Amazon alleged scraped product listings without authorization.
Earlier copyright disputes from publishers including The New York Times, Forbes, and Condé Nast targeted Perplexity's practice of summarizing and reproducing copyrighted content without licensing agreements. Several of those cases remain active.
The data-sharing lawsuit introduces a different kind of risk. Copyright disputes threaten the company's content pipeline. Privacy violations, if proven, threaten its user base.
The Defendant Responses
Perplexity spokesperson Jesse Dwyer said the company had "not been served any lawsuit that matches this description so we are unable to verify its existence or claims." The statement does not address the underlying allegations.
Meta pointed to its advertiser policies, which prohibit advertisers from sending sensitive user information through its tracking tools. That defense, however, addresses what advertisers are supposed to do with Meta's tracking infrastructure, not what Perplexity allegedly did.
Google did not immediately comment.
The gap between Meta's official policy and the complaint's allegations matters. Meta's tracking pixel and SDK are widely embedded across the internet; the tool itself is not unusual. What the lawsuit claims is unusual is that Perplexity embedded these trackers in an AI chatbot interface where users are sharing sensitive, personal queries, and did so without disclosure.
What This Means for AI Search Users
The Perplexity lawsuit arrives at a moment when millions of people are treating AI chatbots as trusted confidants. Users routinely share medical questions, legal dilemmas, financial details, and personal problems with tools like ChatGPT, Claude, Gemini, and Perplexity. The implicit assumption is that those conversations are at least as private as a search engine query.
That assumption may be wrong. Most AI chatbot privacy policies permit some degree of data collection for model training, analytics, or product improvement. But the Perplexity complaint alleges something more aggressive: not just collecting data internally, but transmitting it to third-party advertising platforms in real time.
If the allegations hold up in court, the implications extend beyond Perplexity. Every AI chatbot company will face pressure to disclose exactly what tracking mechanisms are embedded in their platforms, whether third parties receive any user data, and whether privacy or incognito modes actually function as advertised.
For data scientists and engineers who use Perplexity for technical research, the immediate concern is simpler: were your queries about production systems, model architectures, or proprietary datasets forwarded to Meta and Google? The lawsuit suggests they may have been.
If the case is certified as a class action, any Perplexity user could potentially join as a plaintiff. California privacy law provides statutory damages that can scale rapidly across a large user base. Perplexity's next moves in court filings will reveal whether the company disputes the existence of trackers entirely or argues that its data practices were properly disclosed in its privacy policy.
The Broader Pattern
Perplexity's legal exposure is beginning to look systemic rather than incidental. The Amazon scraping injunction, the publisher copyright suits, and now a privacy class action each target a different part of the company's operations, but they share a common thread: allegations that Perplexity built its product by taking things it didn't have permission to take.
Content from publishers. Product data from Amazon. And now, allegedly, the private conversations of its own users.
The company raised significant capital on the promise of building a new kind of search engine, one powered by AI rather than links. That vision attracted users who wanted faster, more conversational answers to their questions. The lawsuit filed Tuesday asks whether those users understood the cost of that convenience.
The Bottom Line
A Utah man trusted an AI chatbot with his financial life. A federal lawsuit now alleges that Perplexity handed those conversations to Meta and Google through hidden trackers that operate even in the platform's Incognito mode.
The case is early. The allegations are unproven. Perplexity says it hasn't been served.
But the questions the suit raises are not going away. If AI chatbots are embedding third-party advertising trackers inside conversations that users believe are private, the entire trust model of the AI assistant industry is built on a foundation that may not hold. The concern echoes the LiteLLM supply chain attack that exposed 95 million monthly downloads to credential theft: when the tools practitioners rely on betray their trust, the damage extends far beyond one company.
Sources
- Perplexity AI Machine Accused of Sharing Data With Meta, Google — Bloomberg, April 1, 2026
- Perplexity AI Sued Over Alleged Data Sharing With Meta and Google — The Decoder, April 1, 2026
- Perplexity AI Machine Accused of Sharing Data With Meta, Google — Claims Journal, April 1, 2026
- Perplexity AI Sued Over Alleged Data Sharing With Meta, Google — NewsBytes, April 1, 2026
- Utah Man Files Class Action Lawsuit Against Perplexity for Sharing Search Data With Google, Meta — Seeking Alpha, April 1, 2026
- Perplexity AI Faces Lawsuit Over Alleged 'Undetectable' Data Tracking Linked to Meta and Google — Storyboard18, April 1, 2026