Neil HodgeMonday, July 28th, 2025
In a ruling filed in June, a senior UK judge warned lawyers that if they rely on fictitious AI-generated cases to present written arguments in court, they could face criminal charges.
Dr. Victoria Sharp, chairman of the King Bench Division of the High Court, said that AI is strong, but widely available generation tools such as ChatGPT cannot conduct reliable legal research, “when artificial intelligence is misused, it has a serious impact on the management of justice and public trust in the judicial system and the judicial system.”
Sitting with Judge Jeremy Johnson in the department court, Sharp was evaluating the role of lawyers following an introduction to two cases where AI tools were suspected to have been used in preparing discussions that were later presented in writing.
in ayinde Case, the lawyer presented five nonexistent cases as evidence. The lawyer said that the generated AI refused to be used to create a list of cases and conducted a general internet search, but could not later identify the source of the fake cases online. In the second case mentioned in the departmental court, Al Haronethe petitioner's lawyers supported his position by submitting a witness statement, including many authorities. The lawyers relied on the research they conducted without the client independently verifying it.
Just because judges appear to have been generous up to now doesn't mean that they're going to be the future.
Melissa Stock
Members of the IBA Business Law International Editorial Board
in ayinde Case, the judge had awarded the legal representative of the claimant and the attorney respectively a wasted fee order to pay the defendant £2,000. The lawyers and lawyers were also referred to the Attorneys Regulatory Authority (SRA). Sharp introduced the barrister to the Bar Standards Committee (BSB). However, the light emptying procedure did not begin. in Al Haronethe lawyer introduced himself to the SRA as well as the departmental courts. Sharp concluded that the light empty threshold was not met.
There is no shortage of specialized guidance available on AI limitations and risks. Bar Council has released guidance entitled early 2024 Considerations when using ChatGPT and generative artificial intelligence software based on large language modelsThis warns that it misleads the court – careless, but still classified as incompetent and grossly negligent, risking disgrace to the occupation, leading to disciplinary action and legal proceedings. Deliberately presenting false material can be considered as overthrowing the court's empt or flow of justice.
Similar warnings are included in the SRA Risk Outlook Report: Use of Artificial Intelligence in the Legal MarketIt was published in 2023 and published in a blog post written by BSB on the theme of ChatGpt in the court.
The AI guidance issued by court and court judicial law and last updated in April states that “all legal representatives are responsible for the materials placed before the court/court and have a professional obligation to ensure that they are accurate and appropriate.”
“AI offers many potential opportunities in terms of how lawyers work, but its use is something that is carefully encouraged. “AI should not compromise on the standards that clients expect from a job quality and confidentiality perspective.”
The so-called “hatography” – fiction generated by AI – problems are not limited to the UK. For example, in Mata v Avianca Inc In 2023, the lawyer prepared the material in the courts generated by ChatGpt. The opposite counsel and the courts were unable to find many of the cases cited. The year after the case, the American Bar Association issued an ethical opinion on lawyer liability using the generator AI.
Chris Howard, co-chair of the IBA Future of Legal Services Committee, said that using AI to ensure cost-effective and rapid results “indicators with accompaniment, with many new cases of malpractice, such as hallucination case citations.” The solution, combined with lawyers who provide reliable and secure features, maintain a strong human surveillance and a full understanding of basic legal research methods, describes it as an “AI research tool.” However, such provisional costs are an issue. ”
It is essential that Paul Marmore, co-chair of the IBA Law Firm Management Committee, has checked the senior lawyer “Loop” who checks work using AI to see work using AI. However, he also added that it would be wise to limit some of the tasks that can be used by very junior members of the team in the litigation context. For example, you might want to create a meeting or note for a few minutes rather than helping with a case file. Furthermore, he adds that output generated to AI should be considered “suggestion only.”
Marmore, head of law firm Shellard, believes that law firms should train all employees to recognize the risks associated with failing AI use. This is especially true with public AI. Law firms risk committing data breach if they enter cases into tools such as CHATGPT that allow data to be made public.
Melissa Stock, IBA member Business Law International The editorial board says there are several lessons to learn from such cases. First, she said, “Judges appear to have been generous up until now doesn't mean that they're going to be the future. Second, “We need to recognize that despite these mistakes, lawyers continue to use AI.”
Stock, a barrister for the London Millennium Chambers, says he can disclose his complaint of using AI during preparation and order his lawyers to disclose where to show it. This will increase transparency in court and enable lawyers to execute casework more quickly and efficiently.
“We need to recognize that AI will increasingly rely on legal professions to perform broader tasks,” she says. “Just because technology is incomplete to date doesn't mean people will stop adopting it. If lawyers at least reveal that they use AI tools and show where they have been applied, they will be able to provide more effective testing and save time.
Image credit: IcedMocha/Adobestock.com
