Overview
- The jury deadlocked 10–2 in favor of the defense on June 26–27, 2026, prompting a judge to declare a mistrial and schedule a retrial for October 19, 2026 with the defendant to remain in custody.
- Prosecutors used the defendant’s ChatGPT history and screen recordings as evidence, pointing to AI-generated images of fire, queries about anger, rants about the wealthy, and a prompt asking whether someone could be blamed if a cigarette started a fire.
- Defense lawyers argued those chatbot interactions were venting or curiosity taken out of context and not proof of planning, a claim that at least some jurors found persuasive.
- The AI records were presented alongside traditional evidence — iPhone location data, security camera footage, and witness testimony — in a federal arson and destruction case that carries up to 45 years in prison if convicted.
- Legal observers say the October retrial could set early precedent on admissibility and interpretation of retained AI chat logs and may prompt changes in platform data-retention policies and how investigators use chatbot records in criminal probes.