A Utah man filed a class-action lawsuit against Perplexity AI on April 1, 2026, alleging that the AI search company secretly transmits users’ complete conversation transcripts to Meta and Google — including sensitive financial, health, and personal data — without disclosing it in its privacy policy.

The case, Doe v. Perplexity AI Inc., was filed in the US District Court for the Northern District of California (case number 3:26-cv-02803). It names Perplexity, Meta, and Google as defendants.

What the Lawsuit Alleges

The 140-page complaint describes a specific mechanism: when a user logs into Perplexity’s website, trackers are downloaded onto their device. Those trackers allegedly capture full conversation transcripts in real time and transmit them to Meta and Google via analytics tools embedded in Perplexity’s interface.

According to the complaint, this happens automatically and continuously — meaning every question a user asks Perplexity, and Perplexity’s full response, is transmitted to Meta and Google before the user sees it.

The data shared is alleged to include:

• Complete prompt text, including questions about health conditions, financial strategies, investment portfolios, and legal matters
• User email addresses for anyone who created a Perplexity account
• Metadata associated with each query

The lawsuit provides a specific example: a user who asked Perplexity “What is the best treatment for liver cancer?” would have had that full prompt transmitted to Meta and Google as a URL string intercepted inside the browser. The complaint notes that family financial information, tax obligations, and investment strategies shared with Perplexity’s chatbot were all captured in the same way.

Critically, the complaint alleges that this tracking continues even when users enable Perplexity’s “Incognito” mode — a setting users would reasonably expect to prevent exactly this kind of data capture.

None of this, the complaint says, is disclosed in Perplexity’s privacy policy.

Perplexity’s Response

Perplexity spokesperson Jesse Dwyer told Bloomberg: “We have not been served any lawsuit that matches this description so we are unable to verify its existence or claims.”

The response is notable in what it doesn’t say. It doesn’t deny the tracking. It doesn’t say the allegations are false. It simply notes that Perplexity hasn’t been formally served — a standard legal posture in the earliest days of litigation.

Meta said the alleged data collection violates its policies for advertisers. Google did not respond to requests for comment.

Why This Matters Beyond Perplexity

This case is about Perplexity specifically, but the question it raises applies to every AI tool your employees use.

When employees use AI assistants to draft documents, research topics, or work through business problems, they typically assume those conversations are private. They are not always right. Analytics trackers, third-party integrations, and vendor data-sharing arrangements vary widely across AI tools, and most end-user privacy policies are written to maximize flexibility for the company, not to give meaningful protection to users.

The Perplexity lawsuit is the first class action to directly target this specific mechanism — the real-time transmission of AI conversation content to third-party advertising platforms — but it is unlikely to be the last. The complaint’s legal argument, that this constitutes a violation of federal and California computer privacy and fraud laws, applies structurally to any AI company running third-party analytics on user conversations without explicit disclosure.

If the lawsuit succeeds, or if it forces a regulatory response, the result could be mandatory disclosure requirements for all AI vendors: a public, auditable list of every third party that receives conversation data, under what conditions, and for what purpose.

That would be a significant structural change to how the AI tool market operates.

What This Means for Business

The AI tools your employees use carry data risk you may not have evaluated. Most enterprise procurement processes treat AI tools like SaaS software — they check the vendor’s privacy policy, perhaps add a data processing agreement, and move on. But the surface area of AI tools is different. Employees share sensitive, unstructured information with AI assistants in ways they would never type into a search engine or a form field. If that data is being captured and transmitted to advertising platforms, the exposure is significant.

“Consumer” AI tools used for business work create enterprise liability. Many employees use personal accounts on tools like Perplexity, ChatGPT, Gemini, or Claude for work tasks — sometimes with sensitive business data. If a class action of this type succeeds, it establishes that users have legal recourse when AI companies misuse their conversation data. That creates downstream liability for businesses that allowed or encouraged employees to use AI tools without evaluating data handling.

Procurement and governance need to catch up. The gap between how fast businesses are adopting AI tools and how rigorously they’re evaluating them is one of the defining risks of the current moment. Asking vendors about data handling — specifically what third parties receive conversation content, under what circumstances, and whether this is opt-out or opt-in — is now a standard due-diligence question, not an advanced one.

The Perplexity lawsuit is in its earliest stage. Class actions of this type typically take 12 to 36 months to resolve, and Perplexity has not been formally served as of April 1. But the filing itself is a signal: AI data privacy litigation has arrived, and the first targets are the consumer AI tools that employees have been using for work.

---

If you want the playbook other teams are using with Claude and Codex right now, grab the free Working With Claude field guide. Download it here.