Choosing a provider
Deleted ChatGPT chats preserved, 20 million logs ordered in a lawsuit
In The New York Times' copyright case, a US court ordered OpenAI in 2025 to preserve deleted ChatGPT chats, and then to hand over 20 million chat logs. Sam Altman himself said that ChatGPT conversations carry no legal confidentiality.
The New York Times sued OpenAI in late 2023, alleging that the company had trained its models on the paper's articles without proper compensation. So that the paper could investigate its copyright claims, in May 2025 the court ordered OpenAI to preserve ChatGPT logs, including those it would otherwise have deleted. According to TechCrunch, this covered the chats of hundreds of millions of users worldwide, with the exception of ChatGPT Enterprise customers. OpenAI called the order an overreach and appealed.
In July 2025 OpenAI's chief executive, Sam Altman, said on a podcast that many people, especially young people, use ChatGPT as a therapist or life coach. As he put it, conversations with a therapist, a lawyer or a doctor are legally privileged, but nothing comparable has yet been worked out for conversations with ChatGPT. TechCrunch added that in a lawsuit OpenAI can currently be required to produce those conversations.
In October 2025 the court eased the order: from 26 September, OpenAI no longer has to preserve all logs, but data already preserved stays, and conversations of accounts flagged by the Times must still be kept.
Meanwhile the court also ordered OpenAI to give the plaintiffs a sample of 20 million chat logs with personally identifiable information removed; the newspapers had originally asked for 120 million. OpenAI objected on user privacy grounds, but on 5 January 2026 US District Judge Sidney Stein upheld the order, finding that users' privacy had been adequately balanced against the relevance of the logs.
What can we learn from it?
Whatever we type into a chatbot stays with the provider: under a court order it may be kept even after deletion, and in a lawsuit it can end up in front of strangers. Professional confidentiality binds the therapist and the lawyer, not the machine. This does not mean AI cannot be used in client work; it means the text should not identify the client. That is why Confiella sends only reviewed, pseudonymised text; under the GDPR pseudonymised text is still personal data. The same practice works when typing by hand: write in general terms instead of names and recognisable details.
Sources
- ABA Journal: ChatGPT creator must turn over 20M chat logs in copyright litigation, federal judge says
- Engadget: OpenAI no longer has to preserve all of its ChatGPT data, with some exceptions
- TechCrunch: Sam Altman warns there's no legal confidentiality when using ChatGPT as a therapist
This article is based on the public sources listed. If you find a mistake, write to us: [email protected]