Hacker News
Australia's Fair Work Commission Orders a Worker to Pay Costs After He Ran His Case Through ChatGPT
Deputy President Michael Easton ruled the week of August 29 in Sadnan Khan's unfair-dismissal claim against ALDI that Khan's AI-generated legal advice was 'plain wrong' and that he had used ChatGPT as a quasi-legal advisor despite repeated warnings his case had no substantial prospects. Easton wrote that 'if Mr Khan had properly read his own AI generated replies … he would have known that his case was doomed,' and ordered $1,230 toward ALDI's costs — a step the tribunal called rare. The failure mode is not hallucination but the litigant answering the wrong legal question and never reading the output closely enough to catch it.
↳ Follow the thread