Florida Murder Suspect Allegedly Used ChatGPT to Research Disposal of Bodies, Prosecutors Say

Hisham Abugharbieh has been charged in connection with the deaths of his roommate and his roommate's girlfriend, both doctoral students from Bangladesh at the University of South Florida, according to court filings reported by The Guardian.

Prosecutors allege that in the days leading up to the students' disappearance, Abugharbieh purchased duct tape and trash bags — items they contend were connected to the alleged crime. Court documents also indicate that he queried ChatGPT, the AI chatbot developed by OpenAI, asking what would happen to a person placed in a garbage bag and thrown into a dumpster.

The alleged use of an AI chatbot to research aspects of a violent crime has drawn significant attention, though legal and technology experts caution against broad conclusions. AI systems like ChatGPT are designed to refuse requests that facilitate harm, and the precise nature of what information, if any, the chatbot provided in response remains unclear from the available court documents.

The two victims, whose names had not been fully disclosed in early reporting, were reported missing before their bodies were subsequently discovered, according to prior coverage from The Guardian. The case has prompted grief and concern among international student communities in Florida.

Abugharbieh has been formally charged with murder. It is unclear from available reports whether he has entered a plea or whether legal representation has commented publicly on the allegations.

The case is ongoing, and prosecutors are expected to present their evidence in court proceedings. No trial date has been publicly confirmed as of the time of publication.

Why This Matters

The alleged use of ChatGPT during what prosecutors describe as pre-meditated murder raises immediate policy questions about whether AI companies have sufficient safeguards to detect and report queries that indicate planning of violent crimes.
The case adds to a growing body of legal precedent in which digital evidence — including AI chat logs — is used in criminal prosecutions, potentially transforming how investigators build cases.
International students, who already face vulnerabilities around housing and social isolation, are at the centre of this case, highlighting welfare concerns for universities hosting large numbers of overseas doctoral candidates.

Background

AI chatbots such as ChatGPT have been commercially available to the public since late 2022, and their use has expanded rapidly across virtually every demographic. OpenAI and rival developers have implemented content moderation systems intended to prevent their tools from providing information that facilitates violence, illegal activity, or self-harm. However, the effectiveness and consistency of these guardrails have been questioned repeatedly by researchers and policymakers.

This is not the first time AI tools have surfaced in criminal investigations. In several prior cases across multiple countries, prosecutors have introduced chat logs, search histories, and AI interactions as digital evidence of intent or pre-meditation. Courts have generally treated such evidence similarly to internet search histories.

The University of South Florida, a large public research institution in Tampa, hosts thousands of international graduate students. The deaths of two Bangladeshi doctoral candidates have reverberated within that community and prompted broader discussion about student safety and housing arrangements.

Key Perspectives

Prosecutors: Allege that the ChatGPT queries, combined with the purchase of duct tape and trash bags, demonstrate premeditation and an attempt to plan the concealment of the crime.

AI and Technology Experts: Likely to note that AI chatbots regularly receive unusual or disturbing queries and are designed to deflect or refuse harmful requests; the key legal question is what the system actually returned in response, and whether that constitutes meaningful assistance.

Civil Liberties and Defence Advocates: May argue that AI query logs, like internet search histories, do not prove intent on their own, and that individuals search for disturbing content for many reasons, including curiosity, fiction writing, or research.

What to Watch

Whether OpenAI is subpoenaed to provide logs of the alleged ChatGPT conversation, and what those logs reveal about the system's responses.
The formal arraignment and any plea entered by Abugharbieh, which will shape the trajectory of the legal proceedings.
Legislative or regulatory responses: any moves by US lawmakers or AI regulators to require AI companies to flag or report queries that suggest imminent violent intent.

ZOTPAPER