It may turn out that all the queries you send to ChatGPT are stored, even if you delete them or request that the artificial intelligence delete them. These queries can be retained and, if necessary, handed over to law enforcement or the courts. This was stated by former Minister of Justice Arpine Hovhannisyan.
NYT vs. OpenAI
According to her, such a decision was made by a New York court in the context of a case that is currently the most high-profile. The case concerns The New York Times vs. OpenAI. The New York Times filed a lawsuit, claiming that approximately 10,000 of its articles were used by OpenAI in violation of copyright.
“This case will have significant, landmark importance for artificial intelligence and copyright issues. However, within the framework of this case, the court made a decision that has had a bombshell effect. The point is that the court ruled that OpenAI is obligated to retain user queries even after they are deleted by users, as they may have evidentiary value. This means serious issues will arise regarding the protection of personal data,” Hovhannisyan noted.
Privacy Under Question
According to her, OpenAI’s director has already announced an intention to appeal this interim decision. However, if the court’s ruling remains in effect, it will have serious consequences:
“A situation may arise where ChatGPT retains all your queries, just as, for example, Google does, even if you delete your history, or like mobile network operators, which store data about your calls and SMS, even if you delete them. In the U.S., within their jurisdiction, this decision will undoubtedly mean that OpenAI will be obligated to provide the data.”
As for jurisdictions outside the U.S., Hovhannisyan says that each case will need to be examined within the framework of litigation, jurisdiction, and procedural rules:
“But one thing is clear: a situation may arise in other countries, including Armenia, where a decision is made to request data from OpenAI in the context of a case. Whether they will provide the data or not, whether they will comply or not, and whether our state’s authority will be sufficient—these are questions that will be answered over time. For now, simply stay informed.”
AI Development and Accessibility at Risk
In addition to the issues highlighted by Hovhannisyan, the NYT vs. OpenAI case could have other long-term and multifaceted consequences, both for AI usage and for copyright and user privacy protection. First, if the court rules that the use of The New York Times’ materials without permission violates copyright, it could limit AI models’ access to published content. AI companies may be forced to filter content, verify licenses, or purchase rights to use texts, images, and other materials. This, in turn, will slow down the pace of artificial intelligence development and increase the cost of AI subscriptions, as companies will have to incur higher expenses to train their models.
Moreover, this case could have international implications. A precedent set in the U.S. could serve as a basis for similar lawsuits in other countries where copyright protection laws also apply. Companies will have to account for the requirements of different jurisdictions, which will ultimately complicate the global use of AI.