In detail
- Prosecution presented defendant Jonathan Rinderknecht's ChatGPT history: requests for fire images, questions about anger, and hypothetical scenarios about fire liability.
- Jury voted 10–2 in favor of the defense; one juror stated she regularly talks to ChatGPT and saw no proof of guilt in its use.
- Judge declared mistrial; case ended without conviction.
Why it matters
This case demonstrates that courts and juries do not yet accept AI interactions as reliable evidence of intent or character—an important boundary for the use of AI data in legal proceedings.
For you Ensure that AI logs in your business are not automatically treated as proof of behavior or intent; instead, document concrete actions and outcomes.