1The Heppner ruling established that consumer AI platform interactions are not confidential. This page documents what that lack of confidentiality is producing: civil litigation against AI companies over how they design their products and what they do with user data.
2This is a distinct category from attorney sanctions. The defendants here are the AI companies, not the lawyers who used them. The theories are product liability, negligence, fraudulent misrepresentation, and unauthorized data disclosure. The exposure is measured in classes of millions and penalties in the billions. Last updated June 4, 2026.
Florida v. OpenAI: The First State Attorney General Lawsuit
3On June 1, 2026, Florida Attorney General James Uthmeier filed an 83-page civil complaint against OpenAI Inc. and its CEO Samuel Altman in the 10th Judicial Circuit of Florida. It is the first lawsuit by a US state against an AI company over the design and safety of its product. It is the first action by any government to seek to hold an AI company CEO personally liable for user harm.
4The complaint alleges OpenAI knowingly released and aggressively marketed ChatGPT to the public, including to children, while concealing serious risks and suppressing internal safety warnings. It names two Florida residents in connection with alleged harms. The complaint alleges OpenAI prioritized speed to market and commercial gain over user safety.
5The legal theories matter beyond Florida. Negligent design, duty to warn, and fraudulent misrepresentation applied to a generative AI product, by a government plaintiff with authority to seek injunctive relief and civil penalties, will shape how courts treat AI product liability nationwide. Attorney General Uthmeier stated he expects other states to follow.
6The questions this complaint asks are the questions every attorney advising an AI deployment client must now answer. What design choices did the vendor make. What warnings did they suppress. Where is the documentation that supervision happened. These are product liability questions applied to software, and Florida filed them first.
Woodard v. OpenAI: The Data-Sharing Class Action
7On May 13, 2026, a federal class action was filed in the Southern District of California alleging that OpenAI routed ChatGPT users' private conversation data to Google and Meta through tracking pixels embedded in the ChatGPT website, without user consent.
8The connection to Heppner is direct. Heppner held that a consumer AI platform is a third party and that data entered into it may not be confidential. The Woodard complaint documents one mechanism by which that data moves: tracking pixels routing user queries to advertising networks. An attorney who used ChatGPT for client-related research during the tracking period may have transmitted client query data to Meta and Google without intending to.
The Discovery Convergence
9These cases converge with the discovery exposure documented elsewhere on this site. In the OpenAI copyright litigation, courts compelled production of 20 million ChatGPT conversation logs and ordered OpenAI to preserve data users believed they had deleted. The product liability suits, the data-sharing class actions, and the discovery orders describe a single reality: data entered into a consumer AI platform persists, moves, and becomes reachable in litigation the user never anticipated.
10Every exposure on this page traces to the same root: confidential information entered a platform the user did not control, governed by terms the user did not negotiate, retained on infrastructure the user cannot reach. A system that runs inside the user's own controlled environment, that does not transmit data to a third-party platform, and that does not route queries to advertising networks, is not subject to these exposures. The difference is architectural.