AI poses an 'existential threat' to the UK immigration system, judges warn
Fake documents, hallucinated submissions and attempts to exploit court procedures could overwhelm legal systems.
Artificial intelligence is an "existential" risk to Britain's immigration appeal system.
That's the warning in a landmark ruling published this month by a three-judge Upper Tribunal panel that identified patterns of suspected fraud involving bank statements, marriage certificates, death certificates and legal submissions that appeared to have been written using generative AI.
The judges identified ten appeals showing characteristics indicative of abuse of the immigration appeals system and examined five in detail in their published ruling.
They warned: "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."
The tribunal said rapid improvements in generative AI were making fabricated documents and images increasingly difficult to distinguish from genuine evidence.
"AI in the modern digital era presents even greater challenges than those perhaps envisaged in 2002," the judges wrote, referring to longstanding legal guidance governing the reliability of documentary evidence.
Fraudulent evidence and AI hallucinations
The panel described a recurring pattern in which applicants submitted limited supporting evidence before their immigration applications were refused, then produced additional documents during appeals.
Sometimes, those documents arrived shortly before a hearing or were submitted without being properly served on the Home Office, preventing officials from verifying their authenticity.
The judges explained: "An overarching characteristic of the applications and appeals we have considered is the steps taken to prevent the Secretary of State for the Home Department (SSHD) from having an opportunity to consider and test the evidence that has been provided, often at the last minute."
READ MORE: NHS warns of “significant” robotics dependency risk as it sets new tech buying rules
The tribunal described cases involving apparently authentic bank statements, contradictory marriage certificates and other documents that were subsequently found unreliable or false.
The judges also identified legal submissions containing hallmarks of generative AI, including overly formal language, rigid formatting, invented case law and incorrect legal citations.
However, they stressed that using AI was not inherently fraudulent and did not establish that the false documents had been generated using the technology.
"The suspected use of an AI generative tool is not by itself an indication of fraud, but the possibility of misuse cannot be underestimated in an increasingly digital era," the judges wrote.
A growing threat to British justice
The judgment follows an earlier warning reported by Machine, in which the Upper Tribunal raised concerns about hallucinated legal authorities submitted by an immigration solicitor.
In that case, the tribunal warned that fabricated legal citations wasted judicial resources and threatened public confidence in the justice system.
The latest ruling raises a broader concern: AI can undermine not only legal arguments but also the evidential foundations on which judges make decisions.
The judges reaffirmed established legal principles requiring applicants to show that the documents they rely on are trustworthy, while instructing courts to scrutinize suspicious evidence and procedural irregularities.
They warned: "The tools needed for creating fake documents are becoming readily available and are rapidly improving in sophistication and can be used by non-technical users."
The ruling highlights a growing challenge for institutions that depend on authentic records: generative AI is making convincing fabrication easier, potentially overwhelming traditional methods of verification.
