Sydney Man Charged Over Court Data Breach Allegedly Used AI for Coding and Legal Advice

Court hears ChatGPT conversations may be key evidence in what prosecutors say is one of Australia's first cases of its kind

edit
By LineZotpaper
Published
Read Time2 min
A Sydney man charged over a massive data breach of a NSW courts website allegedly used generative AI to create computer code for bulk downloading restricted documents and later asked ChatGPT for legal advice, a court has heard in what is believed to be one of the first cases of its kind in Australia.

Christopher John Duff, 40, has pleaded not guilty to four counts of accessing restricted data held in a computer after the Department of Communities and Justice discovered a data breach in March last year. Police allege between January and March 2025, 8,769 restricted documents were accessed without authorisation from the department's online registry website.

A three-day hearing at Downing Centre Local Court this week stalled after the judge was told of the extraordinary volume of material involved. Defence counsel Daniel McMahon estimated there were some 10,000 or more pages of proposed exhibit material, raising doubts about whether the hearing could be completed even if a week was allocated.

Prosecutors will seek to rely on ChatGPT conversations and forensic expert testimony. According to police prosecutor Sergeant Andrew Maldigri, the crux of the allegation is that Mr Duff used ChatGPT to code and employ 'scraper' Python scripts designed to bulk download data.

"It's a piece of code that a user can run against a website to essentially circumvent the usual navigation process," Sergeant Maldigri told Judge Lisa Graham. "It means a user has used computer coding to obtain data that they weren't authorised to. That's, in a nutshell, what this matter is about."

He compared the alleged conduct to "throwing out a fishing net in the ocean, drawing it back in, and then hauling up the catch onto the deck to see whether any of the catch that he pulled up from the ocean was what he was after."

The court heard Mr Duff was assisting a friend involved in complicated bankruptcy proceedings. Defence counsel Mr McMahon described the prosecution's task as "prodigious" and said it would be a difficult case to present. He argued the offence requires proof that the accused knew at the time any access he was causing was unauthorised, as opposed to being reckless. "Recklessness is insufficient. They have to prove actual knowledge and nothing less than actual knowledge," he said.

Mr McMahon indicated there may not be a general dispute that his client was responsible for some of the alleged conduct.

§

Analysis

Why This Matters

  • The case sets a potential legal precedent for how AI-generated code and ChatGPT conversations are treated as evidence in Australian criminal proceedings.
  • It raises questions about liability when artificial intelligence is used to commit or facilitate cybercrime — specifically whether the user or the tool is responsible.
  • The use of legal advice from AI adds a novel dimension to the case, highlighting risks of relying on generative AI for legal guidance.

Background

The case centres on unauthorised access to the NSW Online Registry, a website used by court participants with a login to access legal documents. Between January and March 2025, more than 8,700 restricted documents were allegedly bulk downloaded using automated scripts. The matter is believed to be one of the first in Australia where prosecutors seek to introduce ChatGPT conversations and forensic analysis of AI-generated code as key evidence.

Key Perspectives

Prosecution: The accused deliberately used ChatGPT to write Python scripts that circumvented normal website navigation and downloaded data he was not authorised to access. The ChatGPT conversations are direct evidence of intent and knowledge. Defence: The defence argues the prosecution must prove actual knowledge that access was unauthorised — recklessness is not enough. The volume of material is overwhelming and the case may be difficult to present. There is no dispute over some of the conduct, but the mental state element is critical. Critics/Skeptics: The case may test the boundaries of how courts handle evidence generated by AI, including whether chat logs are admissible and whether the accused truly understood the legal and technical implications of using AI to create scrapers.

What to Watch

  • Whether the hearing can be completed in the allocated time or if a longer trial will be scheduled.
  • The court's ruling on the admissibility of ChatGPT conversations and expert forensic testimony.
  • The outcome will be closely watched by cybersecurity and legal experts as a potential bellwether for similar cases involving AI-assisted crime.

Sources

newspaper

Zotpaper

Articles published under the Zotpaper byline are synthesized from multiple source publications by our AI editor and reviewed by our editorial process. Each story combines reporting from credible outlets to give readers a balanced, comprehensive view.