Perplexity faces class action over sharing chats and data with ad platforms

Perplexity received a class action lawsuit in the USA: the startup is accused of covertly transferring user data to advertising platforms without transparent consent. At the center of the claims are not only technical metadata, but also the very conversations with the chat-bot, where people could discuss finances, law and other sensitive topics.

What is the essence of the lawsuit

An anonymous user who regularly turned to Perplexity for financial and legal advice is the plaintiff. The complaint alleges that advertising tools from third-party companies were embedded within the application, capable of collecting data about a person's actions in the service. If a user was logged in, such tools, according to the lawsuit, could gain access to the original prompts and subsequent messages in the conversation chain.

In essence, the claim is that the content of communication with AI could be used not only to provide answers, but also for ad targeting. The picture in the lawsuit is described even more harshly for guest users. It alleges that third-party platforms had access to the entire chat history with the bot, not just individual technical signals like cookies or interface events.

The plaintiff compares such a scheme to wiretapping: a person talks to a digital assistant expecting privacy, while their requests supposedly are read in parallel by advertising systems. For a product that sells itself as a convenient intelligent interface for search and answers, this is a particularly painful accusation.

What data could be collected

The case materials state that the transmitted data array made it possible to identify a specific person. This is not just about usage statistics, but about a combination of request content, session history, and technical identifiers. Such a combination is especially sensitive in services like Perplexity, because users often formulate requests as personal questions rather than as an impersonal search query. If this data really went to advertising partners, the risks go far beyond typical web analytics.

• original prompts and subsequent messages in the dialogue
• chat history within one or multiple sessions
• personal and technical identifiers associated with the user
• signals suitable for targeted advertising configuration

A separate claim concerns incognito mode. The lawsuit states that even when the feature is enabled, which is supposed to enhance confidentiality, data collection and transmission did not stop. If this is confirmed, the problem will not only be in the fact of tracking itself, but in the discrepancy between user expectations and the product's actual behavior. For AI services, this is critical: people increasingly discuss work documents, deals, medical questions, and personal decisions with bots, assuming at least a basic level of privacy.

Who else is involved in the case

The claims are addressed not only to Perplexity, but also to Google and Meta, whose advertising tools, according to the plaintiff, were used within the application. The logic of the accusation is this: major platforms have their own rules and restrictions for trackers, but in this case they allegedly did not work. The lawsuit also separately criticizes the presentation of the privacy policy. According to the plaintiff, the service did not require explicit consent to the document before use, and the link to the terms was difficult to find in the web version.

The potential scale of the case is large. The class action lawsuit could include users from the USA, whose conversations, it is claimed, were intercepted from the end of 2022 to the beginning of 2026. The compensation amount in the case materials is estimated at more than $5,000 for each individual violation episode, and the total number of such episodes could be in the millions. Even if the court ultimately reduces the amount of claims, for Perplexity this is still a risk of serious reputational damage and expensive litigation against the backdrop of growing attention to AI products.

What does this mean

For the AI market, this is yet another signal: users are willing to share increasingly sensitive information with chat-bots, which means that privacy requirements for such services will become stricter. If the claims against Perplexity are confirmed even in part, the industry will have to explain much more clearly what data is collected, to whom it is transferred, and what private communication mode actually means.