A new study shows AI hallucinations in US courts jumped from 230 to 1,667 cases in one year, with error rates topping 34% at major legal AI platforms.

A lawsuit alleges an AI security company published hallucinated findings, highlighting the growing crisis of false positives threatening cybersecurity

According to the report, documented AI-related filing errors climbed from just 25 cases in early 2025 to 249 by the fourth quarter of that year, with the pace continuing into 2026.

A new study shows AI hallucinations in US courts jumped from 230 to 1,667 cases in one year, with error rates topping 34% at major legal AI platforms.