Nineteenth Lawyer Accused of AI Hallucinations This Week. It Is Tuesday.

A nineteenth lawyer has been accused of including fictitious authorities in a legal submission this week. It is Tuesday.

According to the firm's internal review, the lawyer used an AI tool to identify supporting cases, then verified the resulting citations by asking the same tool whether the cases existed.

It confirmed that they did.

The firm's responsible-use policy requires independent verification. The partner responsible for the matter said he understood independent to mean asking a separate question in a new chat window.

A spokesperson noted that the firm maintained a comprehensive AI policy, compulsory training, an oversight committee and a PDF sent to every fee-earner in February.

“The policy specifically requires lawyers to verify all citations,” the spokesperson said. “We are therefore satisfied that the policy is working.”

The firm's knowledge team has arranged an emergency webinar entitled Does This Case Exist? It is designated advanced training because attendees will be asked to open the case in a source other than the system that invented it.

Several partners objected that the exercise could add time to legal research. Finance has asked whether verification can be completed by a cheaper member of staff, or by a separate AI agent instructed not to hallucinate.

Professional indemnity advisers have suggested that the problem might not be the number of AI policies but the difficulty of persuading people to follow them. The firm intends to add this insight to the next policy update.

14:20 UPDATE: A twentieth lawyer has been identified. His alleged hallucination was not a citation but a statement that his team had completed the required verification. The team says it thought the statement was being checked by the first lawyer.