The rising use of AI will enable parties to falsify evidence to improve their prospects of success. Document forgery predates AI, but traditional verification methods could expose it. AI can now create foolproof false documents that evade those checks.
“The ability of AI models to generate material, including documents, is capable of posing an existential threat to the integrity of the immigration appeal system and requires particular vigilance.” – Hoxha v SoS for the Home Department UT Immigration and Asylum Chambers
The recent decision in Hoxha v SoS illustrates this threat to courts and tribunals. AI was used to manipulate authentic-looking council tax bills, emails and bank statements whose contents did not match the relevant organisations’ records. The Tribunal said that the party relying on a document must establish that it is reliable, and the court must assess reliability in the round, rather than assuming an official-looking document is genuine. Litigators need to establish a document’s provenance, not simply whether it looks genuine, and to distinguish its form from its content.
This article explores two potential consequences for litigation, courts and tribunals:
The increase in complexity, workload and costs.
The potential change of process of challenging falsified documents.
Complexity, workload and costs
“At its simplest we need to differentiate between form and content; that is whether a document is properly issued by the purported author and whether the contents are true.” – Hoxha v SoS for the Home Department UT Immigration and Asylum Chambers
This distinction between form and content is central to assessing a document’s reliability.
As our article explains, identifying false documents can be difficult. Litigators may need native files, metadata or corroborating records to establish how a document was created and whether its content is reliable.
These checks increase review time, workload and costs. Although most documents will not be fraudulent, foolproof AI forgeries may cast doubt on content previously taken at face value.
Challenging false evidence
Our blog also stresses the need to raise suspected falsification with the court promptly. Under Part 32.19 of the Civil Procedure Rules (the CPR), a party is deemed to admit a disclosed document’s authenticity unless it serves a Form N268 Notice to prove documents at trial. Early action is therefore essential.
Yet foolproof AI forgeries may be difficult to detect early, while courts remain critical of late authenticity challenges that should have been raised during disclosure.
Where do we go from here?
The Civil Justice Council’s Final Report 2026 addresses legal representatives’ use of AI in preparing court documents. It recommends amending the CPR to require a specific declaration that AI has not generated or influenced the substantive content of a factual witness’s evidence. However, there have been no proposals or consultations on AI use in evidential documents.
There have been calls to remove CPR 32.19’s presumption of authenticity and require parties to prove each disclosed document’s reliability. Litigators might then need to examine native files, metadata and corroborating records for all disclosed documents, not just those showing signs of fraud. Whether courts would regard those costs as reasonable and proportionate remains to be seen.
An alternative proposed requirement is a disclosure statement confirming that disclosed documents have not been created or amended by AI, reinforcing the contempt-of-court consequences of untrue evidence.
For now, parties should be putting documents to proof as early as possible, seeking native documents, and seeking to engage forensic experts to challenge authenticity where appropriate.
Conclusion
AI’s impact on the court system extends beyond drafting court documents to creating foolproof false evidence. Its growing sophistication may require changes to how courts verify documents and handle authenticity challenges.
For more information, please get in touch with Sarah Jacobs or Abbie Phillpotts.
You can also explore Deception Diaries, our fraud forum providing regular articles with insights on fraud-related issues, legal developments, and practical strategies to help businesses prevent and address fraud.
Our content explained
Every piece of content we create is correct on the date it’s published but please don’t rely on it as legal advice. If you’d like to speak to us about your own legal requirements, please contact one of our expert lawyers.