The Economist’s case study of how AI-empowered individuals are overwhelming a bureaucracy built for the analogue age: Britain’s employment tribunals. Free, AI-powered legal advice should be good news for workers — instead it’s a textbook tragedy of the commons. The canary is “interim relief,” an obscure emergency provision used ~20 times a year across Britain — now ~20 times a month per region, a 100-fold increase. The numbers: claims against employers up 39% in a year (to 50,000), the backlog up 55% to 64,000 cases, a case filed today may not be heard until 2030. “In the age of AI, a system intended to provide access to justice suffers from, if anything, too much access.” The vivid part is the slop: enter a vague grievance into ChatGPT and it drafts a claims form, a “model legal argument,” an “employer defence prediction map” — their fictional claimant bullied for liking horoscopes (not yet a protected belief) was told veganism is. Judges get hundreds of pages citing Magna Carta and hallucinated laws; an NHS employee prepared 67 grievances over 282 pages with Grok and couldn’t say which 10% he’d rely on; a neurodiversity claimant had made 100+ tribunal claims “and yes, he had recently discovered AI.” Tribunals are uniquely vulnerable because they’re a “people’s court” by design — no fee, losing side rarely pays, no penalty for rejecting settlement: “there is no financial mechanism to make parties think hard about the merits.” And the government is pouring fuel on it: the Employment Rights Act adds ~25 more grounds, cuts the unfair-dismissal threshold from 2 years to 6 months, and scraps the compensation cap. The ending reframes it: the slop wave is the temporary phase — specialist AI can already beat human lawyers at drafting, and if it matures, “every worker [gets] the equivalent of a top-flight lawyer in their pocket,” turning a deluge of bad claims into a wave of winning ones. “With AI, power will move faster and further than the politicians imagined.”

Read the full essay at The Economist