Jurors Hear ChatGPT Rebuffed LA Fire Suspect's 2024 Requests for Burning-City Images
Updated
Updated · Bloomberg · Jun 11
Jurors Hear ChatGPT Rebuffed LA Fire Suspect's 2024 Requests for Burning-City Images
3 articles · Updated · Bloomberg · Jun 11
Summary
Federal jurors on Thursday heard that Jonathan Rinderknecht repeatedly asked ChatGPT in 2024 to generate images of a city consumed by flames and floods, months before the Jan. 1, 2025 Palisades Fire.
ChatGPT refused some prompts as too violent, and prosecutors said Rinderknecht cursed at the bot when it pushed back on the scenes he wanted.
ATF agent Michael Montevidoni, who helped retrieve data from Rinderknecht’s two phones, said the requests focused on wealth inequality paired with mass devastation.
The testimony builds on earlier evidence that Rinderknecht used ChatGPT as a "diary" for thousands of questions, including prompts about burning forests, as he faces three arson charges in the blaze that devastated an affluent Los Angeles neighborhood.
His AI chats revealed a fascination with fire. Does that make him guilty of causing LA's most destructive blaze?
A fire smoldered underground for a week before destroying a city. Who is truly to blame for the disaster?
The 2025 Palisades Fire: Catastrophe, Arson Trial, and the Legal Precedent of AI Evidence in Los Angeles
Overview
The report details the trial of Jonathan Rinderknecht, who is accused by prosecutors of intentionally starting the catastrophic Palisades Fire out of revenge against society, leading to the most destructive wildfire in Los Angeles history. The defense counters that Rinderknecht was only on the hill to watch fireworks on New Year’s Eve 2024. The case highlights the clash between the prosecution’s claim of deliberate arson and the defense’s explanation, setting the stage for a high-stakes legal battle over responsibility for the devastating fire.