Federal Judge Rules ChatGPT Records Can Be Used as Evidence in Crypto Fraud Case

Federal Judge Approves Search Warrant for AI Chatbot Records

US District Judge Lorna Schofield in Manhattan has ruled that prosecutors can compel OpenAI to turn over ChatGPT records belonging to Richard Kim, former CEO of cryptocurrency startup Zero Edge, in a securities fraud case 1

1

2

2

. The decision on Monday lifted a temporary block on the search warrant for AI chatbot records, establishing that AI chat logs can be treated as third-party digital evidence subject to legal discovery. Prosecutors allege Kim diverted approximately $3.8 million from a $4.3 million fundraising round for Zero Edge into personal crypto trades and online gambling 1

1

.

Attorney-Client Privilege Does Not Extend to AI Systems

Kim's defense team attempted to block the warrant, arguing that his ChatGPT records contained privileged information related to his case strategy and internal defense tactics. However, prosecutors successfully countered that AI systems are not lawyers and cannot provide the confidential legal counsel required for attorney-client privilege to apply 1

1

. The warrant seeks Kim's OpenAI records from October 2023 through May 2026, including prompts, responses, and account information. Court filings indicate Kim used the AI to generate prompts related to misappropriating investor funds, cryptocurrency trading, and gambling, and later researched his trial strategy following his arrest 1

1

. Kim has pleaded not guilty to securities and wire fraud charges.

Growing Legal Implications of Using AI Tools

This ruling aligns with United States v. Heppner, a landmark decision from earlier this year where US District Judge Jed Rakoff determined that exchanges with Anthropic's Claude chatbot were not protected by attorney-client privilege or work-product protections 1

1

. Rakoff characterized AI platforms as third-party data collectors rather than legal counsel, noting that the defendant used Claude without attorney direction and that privacy terms weakened confidentiality claims. The legal implications of using AI tools continue to evolve as courts establish that AI communications in court can be treated similarly to emails, texts, and search history 1

1

.

Contrasting Rulings Highlight AI Privacy and Discovery Complexities

While federal criminal cases have consistently allowed prosecutors to hand over user records from AI chatbots, a contrasting decision emerged in civil litigation. Justice Rhonda Fischer of Nassau County Supreme Court in New York blocked a subpoena seeking ChatGPT records in Alpha Tech Lending's lawsuit against John Recchio III, finding the chat logs were protected under work product doctrine 3

3

. Fischer cited a Colorado federal court ruling that recognized self-represented litigants' use of AI can resemble "confidential, strategy-laden iterative work product," rejecting what Recchio called a fishing expedition into his private litigation workspace .

What This Means for AI Chatbots Users

The diverging outcomes in criminal versus civil cases signal that legal proceedings involving AI chatbots will depend heavily on context. While Schofield's ruling does not definitively determine whether Kim's ChatGPT records are privileged, it allows the warrant to proceed, meaning the defense cannot prevent OpenAI from complying 2

2

. Kim may still challenge specific records afterward. Prosecutors have indicated they are prepared to respect individual assertions of privilege properly justified by defense counsel after reviewing the data 2

2

. Future cases may establish clearer boundaries around when AI-assisted legal work receives protection, particularly if a chatbot is used under attorney direction. As more people rely on tools like ChatGPT, Gemini, and Claude for research and advice, these rulings serve as a reminder that conversations with AI chatbots can become part of a legal record and are not automatically private or protected.