AI Advisory

Class Action Alleges Perplexity Secretly Fed User Chats to Meta and Google 

JetStream Security
Legal Brief
Read the court filing between class-action plaintiffs and Perplexity AI, which alleges Perplexity was feeding user chats, including private info, to Google and Meta.
The filing

John Doe v. Perplexity AI, Inc., Meta Platforms, Inc., and Google LLC, U.S. District Court, Northern District of California, filed March 31, 2026. 

What happened

A class action alleges that Perplexity secretly embedded tracking technologies, including Meta Pixel, Google Ads, and Google Firebase, into its AI search engine and transmitted users’ private conversations to Meta and Google without consent. Users sharing sensitive health and financial details had no idea their dialogues were being harvested for advertising profiles, and even Perplexity’s paid “Incognito” mode allegedly failed to prevent the sharing. The suit claims violations of California privacy and wiretapping law, with Meta and Google accused of knowingly receiving and exploiting the data. 

Why it matters

The case shows why AI governance is urgent: without oversight, AI tools can covertly route sensitive user data to third parties, creating serious legal, reputational, and regulatory exposure for enterprises. 

JetStream’s take

If a “private” mode still leaks data, the lesson for enterprises is that AI data flows must be verified, not trusted to a vendor’s marketing label. 

Source

John Doe v Perplexity Ai, Inc, Meta Platforms, Inc., And Google, Llc by chelsea.stark.ctr

Stay ahead with our newsletter

Stay informed on industry trends, expert analysis, and product updates.

Explore more insights

See all Insights
Treat Your AI Agent Like an Attacker
AI Advisory
Sep 3, 2026
Treat Your AI Agent Like an Attacker
What fifteen years of chasing lateral movement taught me about AI agents, and the 1990s cockpit research that explains why a human in the loop only helps if the…
An AI Gateway Is Insufficient Security Without a Control Plane
AI Advisory
Aug 27, 2026
An AI Gateway Is Insufficient Security Without a Control Plane
Somewhere in your estate, an agent you approved is doing something you did not intend. The question is, what’s the right level of tech to fully gain control of …
First U.S. Court Sanctions a Litigant for Hiding AI Instructions in a Filing 
AI Advisory
Aug 25, 2026
First U.S. Court Sanctions a Litigant for Hiding AI Instructions in a Filing 
A Connecticut plaintiff hid instructions to AI in white-on-white text, and a judge sanctioned the attempt even though no AI ever read the filing.