OpenAI says a ChatGPT user agreed to the data sharing she is suing over

What OpenAI told the court

OpenAI is asking a federal judge to throw out a privacy lawsuit filed by a ChatGPT user, Amargo Couture. As MediaPost reported, the company filed its latest arguments on Monday with U.S. District Court Judge Dana Sabraw in San Diego.

The core of the defense is consent. OpenAI says Couture was told plainly how her data would be handled and used ChatGPT anyway. In the company's words, that consent "defeats each of her claims."

OpenAI also wants a quick end to the case. It argues that even if everything in the complaint were true, the facts would not show a violation of privacy law, so the case should not go any further.

What Couture alleges

Couture filed her class-action complaint in May. A class action is a lawsuit brought by one person on behalf of a larger group with similar complaints. Here, that group would be other ChatGPT users.

She claims ChatGPT sent data about her online activity to Meta and Google through embedded software, specifically the Meta Pixel and Google's analytics code. These are tools that website operators build into their pages. The complaint says this happened without her real agreement.

She says that starting in 2025 she visited ChatGPT's website and entered queries about her health, her finances and other private information. The suit alleges that information about those sensitive chats ended up with the two companies.

The fight over a buried link

Much of the dispute is about whether OpenAI's notices were visible enough to count. OpenAI points to its privacy policy and cookie notice, which it says state that "prompts and other content" may be shared with vendors and service providers, including analytics providers. It says this applies whatever the subject matter.

Couture's attorneys answered earlier this month that the disclosures were inadequate. They say the link to the privacy policy was tucked at the bottom of a page. They add that it did not look like a link, because it was not in a different color and was marked only by an underline.

OpenAI's Monday filing asks the judge to reject that argument. It notes that Couture does not claim she never saw the notices or that she declined to read the linked policies. It also calls her allegations about consent "conclusory," a legal term for statements that assert something without backing it up with specific facts.

The company adds that users who create accounts agree to OpenAI's terms and are treated as having read the privacy policy.

What to take from it

This is an early stage of the case. The judge has not ruled in the reporting we have, and nothing here has been proven. Couture's claims are allegations, and OpenAI's description of its policies is its own argument.

Still, the exchange shows what may decide cases like this. Courts often ask what a user was told and whether she had a fair chance to notice it. A company can say, as OpenAI does here, that continuing to use a service after being told the rules amounts to agreement. Opponents can say the notice was designed in a way that made it easy to miss.

For ordinary users, the practical point is simple. A chatbot box feels like a private conversation, but the terms behind it can permit sharing with outside service providers. The place to learn what you have agreed to is the policy and the settings, not the tone of the chat.

What comes next

The next step is for Judge Sabraw to decide whether to dismiss the case or let it proceed. If it proceeds, the question of what Couture actually saw and agreed to would probably get more scrutiny.

The outcome could also shape how companies word and display their privacy notices. A ruling that favors OpenAI would strengthen the idea that a linked policy is enough. A ruling for Couture would suggest that how a notice looks on the page can matter as much as what it says.