How did ChatGPT prevent the murder?

The AI security system detected violent content in the user's conversations with ChatGPT and alerted OpenAI, which then notified the FBI.

Has the suspect been convicted?

No, the man has not been convicted yet. The investigation is ongoing, and he is currently under house arrest with electronic monitoring.

What did the defense say about the evidence?

The defense is requesting complete transcripts of the conversations, claiming that only the user's messages were provided, without the AI's responses, which could distort the context.

How many similar cases have occurred in Brazil?

This is the third case in Brazil where an AI system has alerted authorities to a potential planned crime.

Crime

ChatGPT Prevents Murder of 8-Year-Old Boy

B.S. · Prije 2h | 2 min

Brazilian farmer arrested on suspicion of planning to kill his 8-year-old son. AI security system detected the threat and alerted authorities.

Summary

A 36-year-old farmer was arrested on June 19, 2026, on suspicion of planning to kill his 8-year-old son.
ChatGPT's security system detected the threat, and information reached Brazilian authorities via OpenAI and the FBI.
The suspect was released to house arrest with electronic monitoring on August 13, 2026, and the investigation is ongoing.

Make 5MIN.hr a preferred source

An unusual case from Brazil has raised new questions about the role of artificial intelligence in preventing crimes. A 36-year-old Brazilian farmer was arrested on June 19, 2026, in the state of Espírito Santo on suspicion of planning to kill his 8-year-old son, with key information reaching police after an alert linked to his conversations with ChatGPT, as reported by Lika Online.

Chain of Alerts from AI to the FBI

According to Brazilian authorities, the man had been discussing violent intentions with the AI for over two months. His account was flagged by a security system associated with ChatGPT, and OpenAI subsequently contacted the FBI. The U.S. agency then passed the data to Brazil, with information reaching Cyberlab, the technical unit of Brazil's Ministry of Justice, on June 16. The man was arrested three days later.

Motive and Evidence Found

Investigators claim that one motive for the alleged plan was to avoid paying child support, and that the suspect had no relationship with his son. His communications revealed messages related to murder and various violent plans, as well as an interest in the poison ricin. A police search resulted in the discovery of four bottles containing an unknown substance and rope.

Released to House Arrest

After his arrest, the man spent nearly two months in detention. However, on August 13, 2026, he was released to house arrest with electronic monitoring. The court justified this due to the slow investigative process, and the suspect was prohibited from approaching his son and the boy's mother. The investigation against him is still ongoing.

Questions About Privacy and Evidence

The case has raised an important question about how conversations with artificial intelligence are used in criminal proceedings. The defense notes that Brazilian authorities received messages the man sent to ChatGPT, but not the AI's complete responses. Consequently, his lawyers are demanding full transcripts, warning that isolated excerpts without context could lead to misinterpretation of the conversations.

Brazil's Ministry of Justice has described this case as the most serious of its kind in the country to date. This is the third time in Brazil that artificial intelligence has alerted authorities to a potential planned crime. Whether investigators' suspicions will be confirmed and what impact the ChatGPT conversations will have on the ongoing proceedings remains to be seen.