OpenAI Alert Leads to Arrest of Florida Man Accused of Planning Ex-Girlfriend’s Murder

A 25-year-old Florida man was arrested after OpenAI alerted law enforcement to disturbing ChatGPT conversations in which he allegedly described plans to murder his former girlfriend.

Darren Zhou, who worked as a financial analyst at a major investment bank, discussed killing the woman during conversations with ChatGPT in March, according to reports by Futurism and the Palm Beach Post.

Detailed threats raised alarm

Investigators said Zhou's messages went far beyond expressions of anger. He allegedly discussed the woman's hobbies and places she regularly visited, expressed jealousy involving another man, and described a potential murder-suicide scenario.

One plan reportedly involved waiting with flowers in the parking lot of a fitness facility frequented by his former girlfriend. Authorities said Zhou indicated that if she rejected him, he intended to shoot her and then take his own life.

He also allegedly told ChatGPT that he had purchased several firearms and had sent photographs of the weapons to his former girlfriend. His defense attorney later disputed that claim, saying Zhou did not actually possess the guns.

OpenAI contacts authorities

OpenAI detected what it considered a dangerous pattern in the conversations and reported the matter to the FBI, bringing the alleged threat to law enforcement's attention before an attack occurred.

After his arrest, Zhou was released on $100,000 bond and was fired from his investment banking job.

He later pleaded guilty. Under a negotiated resolution, adjudication was withheld and he received eight years of probation, with electronic monitoring required for two years.

Defense points to mental health struggles

Zhou's attorney argued that his client was not a violent or bad person and had been experiencing mental health problems at the time. The defense said his condition has since improved.

The case also highlights growing questions over when artificial intelligence companies should alert authorities after users discuss credible threats of violence with AI systems.

OpenAI previously faced scrutiny after reports that the perpetrator of a mass shooting in British Columbia, Canada, had repeatedly discussed detailed violent scenarios with ChatGPT months before the attack. Families of victims have pursued legal action against the company, while the British Columbia government has also prepared a legal response.

The Florida case shows how conversations with AI platforms can become a serious public-safety issue when threats move from angry rhetoric to detailed planning of real-world violence.