A cautionary statement was issued by a senior judge regarding the potential pitfalls of utilizing artificial intelligence (AI) in preparing legal documents for court proceedings, following the discovery that several references to non-existent legal cases were included in certain documents.

Mr. Justice Sweeting expressed his concerns during a session where he rejected the appeals of two individuals contesting their extradition to Romania for criminal convictions. The Crown Prosecution Service (CPS) had opposed these appeals at the High Court. However, it was revealed by Mr. Justice Sweeting that two legal references cited by the CPS were erroneously generated by an AI and did not have actual existence.

Subsequently, it was disclosed that these two fictitious cases had been included in another document, prompting the matter to be brought to the attention of the CPS prior to a hearing in February. The CPS acknowledged the mistake and issued an apology, clarifying that there was no intention to mislead the court and that investigations were underway to determine the cause of the error.

In a letter from March, the CPS admitted that the inaccuracies in the references were likely a result of utilizing artificial intelligence. Mr. Justice Sweeting highlighted that the crucial oversight occurred when the reviewing lawyer failed to verify the accuracy of the document before submission.

The CPS, in response, emphasized that the error was not deliberate but rather an isolated incident stemming from inadequate verification of the legal authorities cited. Following a comprehensive internal review, which also examined 78 other cases associated with the same lawyer, the CPS concluded that there was minimal risk of a recurrence.

Regarding the impact on the appeal, Mr. Justice Sweeting confirmed that the errors were rectified before the hearing and had no influence on the arguments or court’s decision. He accepted the CPS’s apology and assurance that there was no attempt to deceive.

Looking ahead, Mr. Justice Sweeting noted the increasing use of artificial intelligence in legal processes and the necessity for proper oversight to mitigate risks, particularly in legal research. The CPS, although not utilizing AI in legal decision-making or charging determinations, reaffirmed its commitment to rectifying errors promptly and preventing similar incidents in the future.