×

Mailing List

To help us keep you updated with our handy guides and other useful news, please consider signing up to our mailing list.

It's quick and easy, and we promise not to send you spam or share your details with third parties.

Increasingly, solicitors and Courts alike are coming across more litigants in person who are turning to AI systems such as ChatGPT and Microsoft Copilot to prepare Court claims, witness statements, correspondence and legal submissions in civil claims.

While these tools can appear persuasive, articulate and authoritative, recent Court decisions have demonstrated that reliance upon AI without proper verification carries profound risks.

AI systems do not understand the law in practise in the way a solicitor or Judge does. They predict text based upon patterns in data and whilst these systems can enter individual facts into their data, they cannot analyse and apply those facts in line with the Civil Procedure Rules, case law, and every day practise to the extent a solicitor can.

As a result, AI can produce convincing but entirely false information. These are commonly referred to as ‘hallucinations’.

In legal practice, hallucinations may include:

  • non-existent cases;
  • fabricated quotations from judgments;
  • invented/ incorrect statutory provisions; and
  • invented/ incorrect procedural rules.

A litigant in person may unknowingly file documents containing fictitious case law or inaccurate references to statutory provisions. Even where the mistake is innocent, the consequences can be severe and could include adverse cost orders and damage to credibility.

Furthermore, a litigant in person may rely on AI systems to explain procedural elements of case progression which ultimately, are incorrect, meaning the individual misses a deadline or loses their right to rely on evidence for example, as a direct result.

A number of cases have been heard in the Courts involving AI and the use of falsified (or hallucinated) case law. However, in the recent Decision in Anthony Malcolm Cork & Anor v Mark Smith [2026] EWHC 1199 (Ch), the Judge considered correspondence sent to the court containing AI-generated material, including purported legal references and statutory wording that were inaccurate or entirely fabricated, and the court found that the material had misled the court on more than one occasion.

In this particular case, it was a solicitor from a high end firm that had fallen foul of the AI risk.

Significantly, the Judge had reviewed the material produced by the AI used by the solicitor and said: “on a large number of occasions, it is plainly wrong or, at the very least, extremely misleading.”

The significance of the case extends far beyond professional lawyers. If experienced solicitors can be misled by AI-generated legal material, the danger for unrepresented litigants is even greater.

One of the key principles in English civil procedure is that parties must not mislead the court. Litigants in person are also expected to act honestly and responsibly in proceedings and the risk of falling foul of this is significant.

If you need assistance with your claim, do not hesitate to contact us…..


Published by

Categorised in: , , , ,

Tags: