Brazilian court fined lawyers for hidden prompt injection in a filing

A Brazilian labour court censured and fined two lawyers for hiding white-on-white text in a petition that told any AI reviewing it to respond superficially and not challenge the documents. The court's own AI tool, Galileu, flagged the hidden text; the court said the aim was to bias AI used by the court or the other side, and fined the two firms 10% of the R$842,000 claim, about US$16,000 by the article's estimate. The write-up by Jetstream's general counsel is dated 1 July 2026, so this is an earlier ruling now drawing attention again. On Hacker News (19 top-level comments when we read it) there is no consensus: many said a fine was too light and the lawyers should be disbarred, others were more uneasy that courts use AI to review filings at all, a few doubted the injection would have worked, and one suggested turning outside documents into images and running OCR to strip hidden text. Sorted take: the useful lesson is about your own inbox, not Brazil's courts.