Prosecutors in the trial of Jonathan Rinderknecht — charged with arson linked to one of the deadliest wildfires in Los Angeles history — introduced an unusual form of digital evidence: his ChatGPT conversation logs.
The Case and the Evidence
The Palisades fire began on New Year's Day, 2025, and quickly became a catastrophic event for the region. Rinderknecht faced arson charges, and prosecutors built their case using multiple streams of evidence:
- iPhone location data placing him near the ignition point
- Security camera footage
- Witness testimony
- ChatGPT conversation history
The AI logs proved to be among the most striking elements of the prosecution's argument.
What the ChatGPT Logs Revealed
According to prosecutors, Rinderknecht's ChatGPT sessions included several damning interactions:
- He asked the chatbot to generate images of fire
- He posed the question: "Why am I so angry all the time?"
- He used the AI as an outlet to rant about wealthy people destroying the world
- A screen recording showed him asking ChatGPT whether someone could be held responsible for a fire they lit
These logs painted a picture of a man prosecutors described as emotionally volatile and preoccupied with fire in the days surrounding the incident.
A Mistrial — But a Legal Precedent in the Making
The case ultimately ended in a mistrial, meaning no conviction was reached. But the use of generative AI conversation data as courtroom evidence marks a significant moment in both legal and AI history.
The introduction of ChatGPT logs as prosecutorial evidence raises serious questions about privacy, data retention, and the evidentiary weight juries might assign to AI interactions.
As AI tools become embedded in daily life, this case is unlikely to be the last time a chatbot's conversation history ends up before a judge. Legal teams, privacy advocates, and AI developers alike will be watching closely as courts begin to define how this data can — and cannot — be used.
