People use ChatGPT differently from a search engine. They draft resignation letters, describe symptoms they are too embarrassed to name at a doctor's office, ask for help processing grief, and seek financial guidance they are not ready to discuss with another person. That intimacy is part of what makes the platform useful. It is also what makes a class action lawsuit filed this week in California federal court especially pointed.
The complaint alleges that OpenAI embedded advertising tracking tools on ChatGPT.com: Meta Pixel and Google Analytics. These are not exotic surveillance technologies. They are standard infrastructure deployed across hundreds of millions of commercial websites to measure advertising performance and website traffic. The problem, the lawsuit argues, is that they were never designed with chatbot conversations in mind. Meta Pixel in particular is known to capture form inputs as users type, which on a search engine means queries like "best pizza near me," but on ChatGPT can mean something considerably more sensitive.
The complaint claims that user prompts, email addresses, and personal identifiers may have been automatically transmitted to Meta and Google as a byproduct of this tracking. It alleges violations of the California Invasion of Privacy Act and the Electronic Communications Privacy Act, and seeks both damages and an injunction to halt the practice. OpenAI has not publicly responded to the filing.
The lawsuit names the specific pairing as its central grievance: not just that data moved, but that identifiable users were linked to their chat content. A single prompt about a health condition, attached to a name and email address, can become the basis for a targeted advertising profile that follows someone across the web. That is a materially different privacy risk from an anonymous search query, even a sensitive one.
There is an important tension running through this case that goes beyond OpenAI specifically. Virtually every major web platform has deployed Meta Pixel or Google Analytics at some point. The advertising ecosystem was built on the assumption that websites would share behavioral data with measurement providers, and most privacy policies acknowledge this in language that users rarely read. What the lawsuit challenges is whether that norm is tolerable when the content being tracked includes things people genuinely believe they are saying in confidence.
A similar case was filed against Perplexity AI earlier and later withdrawn, which limits what precedent exists here. The allegations in the ChatGPT complaint are unproven; a lawsuit is an assertion, not a verdict. What the case does accomplish immediately is force a public reckoning with a design question most AI companies have not had to answer aloud: when you embed the advertising plumbing of the commercial web into a product people use like a therapist, what obligation do you have to tell them?
The gap between what a privacy policy technically permits and what a user actually understands when they start typing their symptoms, their worries, or their half-formed confessions into a chat box is real and growing. Search companies have spent years educating users that their queries are used for advertising. AI chatbot companies have not had that conversation yet. This lawsuit may be the beginning of it.