A group of lawyers has been fined $12,000 after a federal judge ruled they submitted court filings containing AI-generated “hallucinated” material.
Artificial intelligence tools such as ChatGPT, Grok, and Copilot are increasingly used in workplaces to speed up routine tasks, but their use has repeatedly caused problems in the legal system. Courts have seen multiple cases where AI-generated legal citations and reasoning turned out to be fabricated.
That issue surfaced again in Kansas, where U.S. District Judge Julie Robinson imposed fines after lawyers relied on unverified AI-generated research in court documents. In her ruling, Robinson said competent attorneys should be fully aware of the risks of using generative AI for legal research without independently verifying its accuracy.
One lawyer, Sandeep Seth, received the largest fine of $5,000 and described the incident as an “embarrassing lesson.” Kenneth Kula and Christopher Joe were each fined $3,000, while another attorney, David Cooper, was fined $1,000 over faulty citations.
Judge Robinson noted that the growing number of cases involving fabricated legal authority generated by AI is “staggering,” highlighting the broader risks of relying on unverified AI tools in legal proceedings.