British employment tribunals and US community colleges face two different pressures as AI makes claims and coursework cheaper to produce at institutional scale.
Artificial Intelligence··Morning
A wave of claims reaches the tribunal
Workers in Britain are using ChatGPT or Grok to write employment claims instead of paying lawyers, The Decoder reports, citing The Economist. In the year to March 2026, claims rose 39 per cent and the backlog grew 55 per cent to 64,000 unresolved cases. A memo by tribunal presidents Barry Clarke and Susan Walker describes filings hundreds of pages long, filled with fabricated laws and unrealistic demands. Applications for interim relief have increased a hundredfold. Generative AI is one part of the pressure rather than the only cause. Labour's Employment Rights Act adds about 25 new grounds for claims and removes compensation caps, a separate change that may also affect volume. The report does not suggest that every AI-written claim lacks merit. It shows that once producing a long filing becomes cheap, tribunal staff must separate fabricated references and impractical demands from genuine disputes within a larger stream. The Decoder also relays a contrasting study from Pakistan: judges given AI tools and training resolved more cases faster. The same technology can increase intake on one side while, with appropriate tools and training inside the institution, helping cases move more quickly on the other.[1]
The real payment behind a fake student
US community colleges are encountering pressure through a different entry point. Fraudsters enrol fake students in courses, collect financial aid and use AI to complete the coursework needed to keep those students looking active, The Decoder reports, citing The New Yorker. The abuse concentrates in asynchronous online courses, where students can remain anonymous. David Song, a professor at East Los Angeles College, says he noticed the pattern several years ago when students with generic Anglo-Saxon names appeared in his history course at a college whose student body is mostly Latino and Asian. Song permits AI in his courses but requires students to disclose its use; he says that rule goes unobserved even when content is plainly generated. History professor David Roach estimates that more than half of his students use AI for papers. The institutional problem extends beyond poor coursework. A false identity, an application for financial aid and automatically produced assignments become parts of one fraud operation. The anonymity and flexible timing that make online courses accessible to genuine students can also let fraudsters sustain many fake enrolments with less human labour, leaving colleges to distinguish participation from an automated appearance of participation.[2]
Cheaper production raises verification costs
The shared pressure in these cases begins with AI reducing the cost of producing material submitted to an institution. A long tribunal claim and a term's worth of community-college assignments can both be created with less effort. The institution's decision work remains: a tribunal must distinguish valid legal grounds and workable demands, while a college must identify a real student and meaningful participation. That similarity does not make the abuses identical. In Britain, genuine people exercising a right to file appear alongside fabricated laws and a growing caseload; in the US college story, false identities and the aim of taking financial aid establish a fraud operation. The common consequence is that the volume of documents or coursework becomes a weaker signal of human intent and validity. The study involving judges in Pakistan also shows that an institutional response can include trained internal use of AI to process cases faster. Together, the reports make stronger identity, source and content verification visible as part of preserving access. Faster production creates volume at the entrance, while public employees still have to determine the legal, educational or financial meaning of each submission. The workload shifts toward deciding which material represents a real person, a sound claim or genuine participation.[1], [2]
