ChatGPT logs failed to prove arson — will courts ban AI character evidence?
prosecutors' use of chatgpt logs to establish motive in the palisades fire arson trial backfired, resulting in a hung jury. this sets a skeptical precedent for admitting ai chat histories as character evidence in criminal trials.
in a landmark case for ai in the legal system, prosecutors used jonathan rinderknecht's chatgpt logs—which included requests for fire images and rants against the wealthy—to argue he possessed the motive and mindset to commit arson. the defense countered that the logs were benign, and jurors ultimately agreed, with one noting that using chatgpt does not indicate a character flaw. the 10-2 vote in favor of the defense forced a judge to declare a mistrial. this outcome suggests that courts may struggle to convince juries that ai chat histories reflect criminal intent rather than casual curiosity or role-play.