AivexaNewsSearch
AI news for builders and product teamsChecked every hour

“thegentlemanpervert” producing AI CSAM is a NJ prosecutor, say cops

Collected Oct 16, 2026

Brian Richard Fitzgerald, 51, New Jersey's deputy attorney general, was arrested at his home on Thursday on felony charges related to AI-generated child sexual abuse material, or CSAM, according to Bucks County Detectives. New Jersey officials accuse him of using AI to turn photos taken from social media of real minors into sexualized images and then distributing those images on social media.

The charges are possession of CSAM, dissemination of photos or film of CSAM, and criminal use of a communication facility. The case began when the National Center for Missing and Exploited Children received a tip in July that a Discord user had shared an illicit image. Detectives traced the account using what the press release described as digital forensics and subscriber records linked to the Discord account and associated Internet protocol addresses, which led them to Fitzgerald's home address.

A search on Thursday turned up prompts and scripts used to create artificially generated CSAM with an AI program, according to the press release. Investigators also reported finding Discord chat logs in which Fitzgerald, under the username "thegentlemanpervert_2160_07833," allegedly admitted to making multiple illicit images with artificial intelligence and sharing them with other users.

An account presented as Fitzgerald's LinkedIn shows he had been a deputy attorney general since 2016, preparing insurance fraud and insurance licensee cases for trial as a state prosecutor. Joe Khan, identified as New Jersey's district attorney, said in a statement that the state aligns with the Department of Justice in treating AI CSAM as equally harmful as CSAM. "No one is above the law, least of all someone entrusted with upholding it," Khan said, adding that the office will pursue such cases aggressively and hold offenders accountable regardless of profession or title.

The arrest is unusual because the accused held a senior post in a state attorney general's office, the legal and law enforcement arm of state government. The source notes that other law enforcement personnel have faced similar accusations: earlier this year a Pennsylvania police officer was arrested over AI-generated porn made from driver's license photos, and more recently in Korea a police officer was arrested for producing AI porn that included CSAM, according to The Korea Times.

Federal authorities have treated AI-generated CSAM as covered by existing child exploitation law since at least 2024, when the DOJ brought one of the earliest arrests. That case charged a Wisconsin man with spreading thousands of realistic images of prepubescent minors across Telegram and Instagram, and the department then clarified to law enforcement that material generated by AI still counts as CSAM. For developers, that guidance matters because the tools involved are ordinary text and image generators; the offense turns on the content and its distribution, not on whether a real child appears in the file.

The mechanics described here are typical of how these investigations move: a platform report flows to the National Center for Missing and Exploited Children, which passes tips to police, who then use subscriber records and IP logs to tie an account to a physical address. Prompts and scripts recovered from a device can function as evidence of intent and of the generation process itself. This suggests investigators are increasingly comfortable treating prompt logs and chat histories as core case material rather than incidental artifacts.

Several things remain unresolved in the public record: Fitzgerald has been charged, not convicted, and the reporting does not describe a plea or a defense response. The precise charges, potential penalties and court dates are not laid out here. A likely trade-off for platforms is that automated detection systems built around known CSAM hashes may not flag wholly synthetic images, pushing trust and safety teams toward classifier-based detection and toward logging that can support later subpoenas. My own reading is that cases like this will keep testing how existing statutes apply to generated media, and that the practical burden will fall on the services where such material is shared to detect it and preserve the records police need.

Why it matters: this case shows that AI-generated CSAM is being prosecuted as CSAM, and that investigators will build cases from Discord reports, IP records and recovered prompts. For product and platform teams, that means content policies, detection and data-retention practices need to treat synthetic sexual imagery of minors as a law enforcement matter, not just a moderation edge case. Developers building image or chat features should assume prompts, logs and account records can become evidence, and that the legal exposure does not depend on whether a real child was photographed. Inference: expect continued pressure on platforms to report and preserve such material, and closer scrutiny of AI tools whose outputs can be repurposed this way.

Read at Ars Technica · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

Discord user flagged AI CSAM, triggering arrest of New Jersey deputy AG.