Israeli Court Rules AI Use Doesn't Absolve Litigants of Responsibility

The Tel Aviv District Court, acting as an administrative court, has rejected an appeal filed by Russian citizen Kristina Bril, which was drafted with the assistance of ChatGPT. The court stated that litigants using artificial intelligence remain fully responsible for the content of their submitted documents, even if they lack legal representation. Bril had sought asylum in Israel, claiming persecution in her home country due to her opposition views and support for Ukraine. Her asylum request was denied by the Population and Immigration Authority, and the Immigration Appeals Tribunal refused to halt her deportation pending her appeal. Bril, who does not speak Hebrew or English, filed her appeal independently, noting that the text was generated and translated by ChatGPT and asking the court to consider its substance and overlook any errors in wording or translation. Judge Oded Maor observed that while the filing did not contain fabricated precedents, it was incoherent and imprecise, requiring him to reconstruct some facts from accompanying documents. He ultimately dismissed the appeal on its merits, partly based on a recording of an interview with Bril, which suggested she had not faced persecution and had considered returning to Russia. The judge dedicated a significant portion of his ruling to the use of AI in legal filings. He noted an increasing number of self-represented litigants using AI to draft documents and including disclaimers about potential inaccuracies, requesting leniency for errors. Judge Maor affirmed that using AI for drafting and translation is permissible if the user understands and verifies the output, but emphasized that access to justice does not negate responsibility. He explained that a disclaimer merely warns of potential harm rather than preventing it, shifting the burden of verifying unverified documents to the opposing party and the court. He also found the tribunal's requirement for Bril to submit an affidavit confirming her understanding and verification of the AI-generated documents to be lawful. Judge Maor recommended that Bril and others in similar situations seek state legal aid or utilize legal clinics.