CPS Prosecutor Cites Imaginary Legal Precedent Generated by AI
A senior prosecutor's reliance on AI-hallucinated case law sparks urgent justice system concerns.
1AI Hallucination in Court
A senior Crown Prosecution Service lawyer, Ravinder Dhillon, cited a nonexistent legal precedent during a high court case regarding the prosecution of a child. The citation, labeled CPS – R v J [2024] EWCA Crim 3529, was entirely fabricated, highlighting the growing risk of AI-generated misinformation infiltrating formal judicial proceedings.
2Origin of the Error
The erroneous reference originated from an AI-generated summary obtained via an internet search. Although the CPS maintains a policy requiring human verification for legal research, the prosecutor failed to authenticate the source, leading to the submission of a fictitious case to the court during a sensitive child rape prosecution dispute.
3Institutional Response Failure
Two senior high court judges described the explanation for the error as unsatisfactory. Legal experts labeled the incident a serious failure, noting that when state representatives rely on unverified AI outputs, it undermines the fundamental norms of trust and authority that underpin the entire legal system in England and Wales.
4Rising Frequency of Errors
This incident marks the second known instance involving the CPS, following a February case where a lawyer cited nonexistent authorities in an extradition matter. Currently, a tracker maintained by barrister Matthew Lee lists 101 confirmed or suspected AI-related errors in UK courts, suggesting that these public cases represent only the tip of the iceberg.
5Broader Judicial Warnings
In June 2025, the high court urged the Bar Council and Law Society to address AI misuse as a matter of urgency. Dame Victoria Sharp warned that such errors carry serious implications for public confidence, noting that lawyers misusing AI could face severe sanctions, including contempt of court proceedings or police referrals.
6CPS Policy and Mitigation
The CPS stated it has a strict policy limiting AI use to ensure it does not influence legal decisions. Following the incident, the agency conducted an internal review and shared findings with staff, emphasizing that all case law must be cited exclusively from trusted, verified legal databases rather than general internet searches.
7Privacy and Operational Risks
Beyond the risk of false citations, experts warn of significant privacy and confidentiality dangers when using AI tools with sensitive legal data. The reliance on AI is also creating additional labor for barristers, who must now dedicate extra time to verifying the authenticity of every legal authority presented in case papers.
8Implications for Legal Integrity
The incident underscores the tension between efficiency-seeking AI adoption and the rigorous standards of legal practice. As the state wields immense power, the use of hallucinated precedents threatens the legitimacy of judicial outcomes, necessitating a shift toward more robust human-in-the-loop verification processes to preserve the integrity of the justice system.