ChatGPT Is a Huge Snitch - New York Magazine

Photo-Illustration: Intelligencer; Photo: Getty Images In 1998, government lawyer Stephen Houck made a novel argument . Asked why the U.S. wasn’t summoning Bill Gates to testify in its antitrust case against Microsoft, Houck said, “Because we have his email.” The New York Times called it “the first major email trial,” in which messages that were voluminous, discoverable, and “alive with ideas and competitive zeal, punctuated with profanity and exclamation points” featured heavily. Soon, there were widespread official policies — not always heeded — about what could or should be said in messages written on the computer. In 2005, a computer consultant was convicted of killing his wife after prosecutors presented a search history that included terms like “neck,” “snap,” and “break.” In the press, it was dubbed the “ Google Murder ” — like the “email trial,” a harbinger of countless legal proceedings and news stories that would surface digital communications that were mistaken for private. Now, interactions with AI are finding their way into the public record. 404 Media reports : An expert witness testifying in a lawsuit about liability for a Houston explosion that killed three people and destroyed roughly 200 homes used ChatGPT to write significant portions of his “expert report.” [His prompts] showed that he asked ChatGPT to help him “create an exceptional expert witness report defending the standard of care at 3M,” and that the report should “show how 3M is 0% at fault for the explosion at Watson Grinding.” Juicing an LLM for “expert testimony” to defend a chemical company is an extreme example of a new phenomenon. But chatbot logs have been leaking into civil and criminal legal proceedings for a while now. In 2024, a Virginia man was convicted of murder with the help of a chat transcript that captured his interactions with Snap’s AI bot. (Snap’s bot had told him that “violence is never the answer.”) In 2025, it was revealed that the man accused of setting the Palisades Fire had asked ChatGPT, “Are you at fault if a fire is lift [ sic ] because of your cigarettes?” In 2026, a Missouri teenager might have gotten away with vandalizing a bunch of cars at MSU had he not described his crime to ChatGPT and asked , “Is there any way they could know it was me?” Chat logs have been used to incriminate AI companies themselves after their bots have been accused of encouraging crimes or self-harm. More similar to the 3M example is a case in which the CEO of a South Korean video-game company was caught conspiring with ChatGPT to avoid paying out executives of an acquired firm before firing them. (The chat logs helped lead to their eventual reinstatement.) Recovered chat logs have also become a feature of government inquiries around the world: In the U.K., technology secretary Peter Kyle was revealed to have asked ChatGPT in 2025 to explain why AI adoption seemed slow in his country; in 2024, shortly before South Korean president Yoon Suk Yeol declared martial law, his head of security was using ChatGPT to learn about coups. Should Democrats take control of the House in November, we can expect to hear a lot about our own government’s use of AI, revealed in the form of presumed-to-be-private chat logs. (This, I’ll bet, is how we’ll learn a lot more about DOGE.) Email transferred a whole lot of previously written and spoken communication into a new, searchable, and recoverable format, followed, eventually, by workplace instant-messaging and collaboration apps. It also induced new sorts of communication that exploded in volume: mass emails, long threads and chains, summaries of meetings, etc. In contrast, nonsocial tools like search engines seemed to create an impression of privacy, leading people to confide incredibly intimate and potentially incriminating information into text boxes administered by companies like Google or by their employers, much of which could be reproduced in court or in public. Chatbots are a fusion of the two and, in most cases, legally identical: a private text box that encourages extensive conversation, sycophantically holding users’ hands as they talk themselves into future liability. You’re led to believe you’re talking to a little guy in the computer; in fact, as with most computer and smartphone interactions these days, you’re also filing text into a searchable database. Add in the rise of AI-powered meeting transcription and social surveillance , and the range of materials generated in the course of a regular computer-job workday will soon be breathtaking. At the least, this should make future white-collar prosecutions a little easier — that is, if the government gets around to caring about those again . Tags: hard paywall john herrman screen time technology artificial intelligence ChatGPT Is a Huge Snitch Your product is saved! You’ll receive emails when your saved products go on sale. Manage preferences.