It’s been a busy year in the field of AI and the law, with UK and US court decisions showing how courts are adapting to deal with the new legal challenges posed by AI. This article summarizes some of this year’s most notable AI law developments.
fake quotes
The potential misuse of AI has recently come under scrutiny in courts, where lawyers and barristers were found to have cited fake cases in their arguments, witness statements and applications as a result of the use of AI. This has led to public criticism, wasteful cost orders and referrals of those involved to professional regulators. There have been two recent court-directed cases in the UK.
- in the case of Ayinde R (pending) v Qatar National Bank QPSC & Anor [2025] EWHC 1383 (Administrator) The defense team cited five sham lawsuits, but the barrister was unable to provide a proper explanation of how this happened, even when challenged. The Head of the Crown Bench Division has provided important guidance on the use of AI in court proceedings, given the professional and ethical responsibilities and obligations of lawyers in court, and called on regulators to urgently consider whether further action should be taken.
- in the case of D (child) (not attending school) The mother submitted a “long” brief on behalf of herself, including several incorrect and non-existent quotations, which she acknowledged were generated with the aid of artificial intelligence. The Court of Appeal found that the mother did not intend to mislead the court, and the judges, while sympathetic, stressed that all parties have a responsibility to ensure that the incidents cited are genuine and provide valid legal authority.
These cases serve as a reminder of the importance of using AI properly and constantly checking all outputs. The courts have made it clear that the onus is on the individual to ensure that the information submitted to the court is true and accurate.
Latest judicial guidance on the use of AI
The guidance, originally published in December 2024, has been updated twice this year. The October 2025 guidance adds a glossary of common terms and expands on the risk of bias in training data and the risk of AI illusions producing inaccurate or misleading information. It provides further advice on confidentiality, reminds judicial officials not to enter personal information into public AI tools, and tells where to report inadvertent disclosures as data incidents. Lord Justice Barth, Chief Judge for Artificial Intelligence, said: “The use of AI by judicial bodies must be consistent with their overriding obligations to protect the integrity of the administration of justice and uphold the rule of law.”
Lawyers Council guidance on AI
On November 27, 2025, the Bar Council published updated guidance on the use of ChatGPT and GenAI. The report concludes that while there is nothing inherently inappropriate about using trusted AI tools to support the delivery of legal services, these tools need to be properly understood and used responsibly by individual practitioners. Our Guidance Hub highlights the key risks associated with LLM. Hallucination; information disorder. Bias in data training. Training on mistakes and sensitive data.
Getty Images v Stability AI
The UK High Court has handed down a mixed judgment in what is widely seen as the first victory for the AI industry. The main copyright infringement claim was dismissed because the training of the AI model took place outside the UK, rendering the claim devoid of any regional basis. Regarding secondary copyright infringement, the judge ruled that while an “article” may constitute an intangible object, an AI model such as Stable Diffusion that does not and has never stored a copyrighted work cannot constitute an “infringing copy.” Therefore, Getty’s claim failed. Some issues have not been tested in court because Getty has withdrawn some of its claims. More cases are expected in the region.
Unitel Direct Ltd v Racing Edge Auto Repairs Ltd & Ors 2025 EWCC 3 (CC)
Unitel Direct Limited has also brought a number of claims in the English courts against various defendants for alleged unpaid fees purported to be oral corporate agreements relating to online advertising. The legal issue concerned the formation of a contract, and Unitel relied on a key part of its evidence, a transcript of a phone conversation allegedly produced using a third-party AI-driven transcription service. However, the judge questioned the reliability of these records due to various inconsistencies and therefore gave limited weight to the evidence. The judge indicated that the court would require a solid foundation. – Complete audio recording, proven accuracy, consistent timestamps – Before we put a lot of emphasis on AI-generated documentation. This highlights the growing need for standards to verify evidence produced by AI systems in litigation.
Courtroom boundaries: AI avatars not welcome
Courts have set clear limits on how AI can be used in legal proceedings. A recent case in the US has set an important precedent for the use of AI in court proceedings. A lawyer tried to use an AI avatar to argue his case in a New York courtroom, but a judge flatly rejected the approach. The court’s decision highlights that while AI is transforming many aspects of legal practice, fundamental boundaries still exist in court proceedings that require human presence and accountability. The justices emphasized concerns about authentication, responsibility for the arguments presented, and the fundamental nature of court proceedings as human interactions. These concerns are reflected in the steps courts and professional standards bodies are taking to clarify when and how AI can be appropriately used in legal practice.
On the front lines of algorithmic liability: The UnitedHealth case
AI-driven decision-making systems create new liability risks, especially when they override human judgment. Another lawsuit in the US – class action lawsuit against UnitedHealth Group and other insurance companies (Estate of Gene B. Locken et al. v UnitedHealth Group, Inc. et al. al) – illustrates the increasing legal risks surrounding algorithmic decision-making in healthcare. Multiple class action lawsuits allege that insurance companies have deployed AI to override doctors’ decisions and deny coverage, despite documented high error rates in algorithmic assessments of patient needs. plaintiff – On behalf of patients whose insurance has been terminated and their assets, we argue that the insurance company’s AI-based denial of insurance claims amounts to breach of contract, breach of good faith, violation of fair trade, unjust enrichment, and bad faith in insurance. Defendant denies these allegations. This case highlights the “black box” problem of AI. Even developers may not be able to explain why their algorithms made the recommendations they made, creating new challenges for both litigation and compliance strategies.
In general, we expect to see more commercial disputes related to technology-enabled fraud, particularly where AI is at the center of disputes. We also expect more copyright and trademark issues to be heard in court in the wake of the Getty case. As evidenced by some of the cases mentioned above, such disputes acquire new complexities as they challenge the application of the usual elements of a cause of action and will require both parties and judges to grapple with applying established legal doctrine to new situations.
