An Oregon federal court imposed $110,000 in sanctions on a law firm after associates used an AI tool to draft briefs containing fabricated case citations — and partners failed to supervise or verify the research. The penalty, one of the largest on record for AI-related misconduct, establishes that supervisory partners bear full responsibility for AI errors made by junior staff.
A New York federal judge sanctioned attorneys from Mata v. Avianca after they cited six entirely fictitious ChatGPT-generated cases in a court filing — and then doubled down when challenged. The 2023 ruling by Judge Castel is the defining global precedent on AI hallucinations in legal practice, imposing financial penalties and requiring corrective filings from all counsel involved.
A comprehensive roundup of every major Indian court and tribunal ruling on AI hallucinations in legal proceedings, covering the Supreme Court, Bombay HC, Delhi HC, AP HC, and ITAT Bangalore. Covers the pattern of sanctions, the Bar Council referrals, and what Indian lawyers must do right now to ensure AI-assisted research does not expose them to professional discipline.
The Income Tax Appellate Tribunal, Bangalore Bench, allowed the Revenue to withdraw its own order after conceding it had relied on AI-generated case citations that could not be verified. The episode — a tax authority undoing its own decision because of AI hallucinations — is a rare instance of a government body acknowledging and correcting AI-assisted legal error.
The Bombay High Court awarded adverse costs against a litigant after counsel submitted a legal brief containing multiple AI-hallucinated citations — cases that simply did not exist. The court held that relying on fabricated precedent without verification constituted a breach of professional duty, and that judges’ time is too valuable to be wasted correcting lawyers’ AI-generated errors.
The Delhi High Court dismissed a challenge to a GST show-cause notice, then discovered the petitioner’s counsel had cited AI-generated fake judgments throughout the pleadings. The court imposed exemplary costs and referred the matter to the Bar Council of Delhi, signalling that AI-assisted fabrication of case law is professional misconduct warranting disciplinary consequences beyond mere dismissal.
The Bombay High Court quashed a ₹2.84 crore income tax reassessment order after discovering the Revenue’s legal brief cited judgments that did not exist — fabricated by an AI tool. The court held that proceedings built on non-existent precedent were fundamentally flawed and could not stand, marking one of India’s first formal judicial rejections of AI-hallucinated legal authority.
India’s Supreme Court and Andhra Pradesh High Court became early AI law landmarks when advocates submitted fake judgments generated by artificial intelligence. The Supreme Court flagged the fabricated citation in a major criminal appeal; the AP HC separately imposed costs on a lawyer who tendered AI-hallucinated cases. Both decisions send a clear warning: Indian courts will not tolerate AI-generated misinformation.
