概要: 報道によると、ジュリアン・ニールズ連邦判事(ニュージャージー州民主党)とヘンリー・ウィンゲート連邦判事(ミシシッピ州サウスダコタ州)は、両連邦裁判所のスタッフがAIツール「ChatGPT」と「Perplexity」を用いて、捏造されたとされる引用、事件、当事者を含む判決文を起草したことを認めた。これらの誤った提出書類は審査前に記録に残されたと報じられており、上院と司法当局による調査が行われた。両判決は撤回された。
Editor Notes: Timeline notes: The reported timeline of events is as follows: (1) 06/30/2025, U.S. District Judge Julien Neals (District of New Jersey) issued an opinion in a securities case against CorMedix Inc. that reportedly contained fabricated case citations and quotes later traced to ChatGPT use by a law school intern; (2) 07/20/2025, U.S. District Judge Henry Wingate (Southern District of Mississippi) issued a temporary restraining order with fictitious plaintiffs and quotations, reportedly drafted with Perplexity AI by a law clerk; (3) 07/22/2025, Wingate retracted and reissued the order after being alerted to errors; (4) 10/20–10/21/2025, both judges admitted AI involvement in letters to Senator Chuck Grassley and the Administrative Office of the U.S. Courts; (5) 10/23/2025, Bloomberg Law reported the admissions; (6) 10/29/2025, The Washington Post published a detailed account confirming Senate inquiry and judicial reforms. The incident ID was added on 10/29/2025.
Alleged: OpenAI と Perplexity AI developed an AI system deployed by United States District Court for the District of New Jersey , United States District Court for the Southern District of Mississippi , Julien Neals , Henry Wingate , Unnamed law clerk (S.D. Miss.) と Unnamed law school intern (D.N.J.), which harmed United States District Court for the District of New Jersey , United States District Court for the Southern District of Mississippi , Julien Neals , Henry Wingate , Plaintiffs and defendants in CorMedix Securities Litigation , Plaintiffs and defendants in Jackson Federation of Teachers et al. v. Mississippi State Board of Education , Epistemic integrity と Judicial integrity.
関与が疑われるAIシステム: ChatGPT と Perplexity
インシデントのステータス
Risk Subdomain
A further 23 subdomains create an accessible and understandable classification of hazards and harms associated with AI
3.1. False or misleading information
Risk Domain
The Domain Taxonomy of AI Risks classifies risks into seven AI risk domains: (1) Discrimination & toxicity, (2) Privacy & security, (3) Misinformation, (4) Malicious actors & misuse, (5) Human-computer interaction, (6) Socioeconomic & environmental harms, and (7) AI system safety, failures & limitations.
- Misinformation
Entity
Which, if any, entity is presented as the main cause of the risk
AI
Timing
The stage in the AI lifecycle at which the risk is presented as occurring
Post-deployment
Intent
Whether the risk is presented as occurring as an expected or unexpected outcome from pursuing a goal
Unintentional
インシデントレポート
レポートタイムライン
Loading...
2人の連邦判事は、職員による人工知能(AI)ツールの使用が誤った判決の原因であると非難し、自らの名義で発行された文書を職員がどの程度精査しているのか疑問を呈した。
ニュージャージー州のジュリアン・ニールズ とミシシッピ州のヘンリー・ウィンゲート の連邦地方判事は、連邦裁判所行政局宛ての書簡の中で、AIの欠陥を認めた。上院司法委員会のチャック・グラスリー委員長(アイオワ州共和党)の質問に応えて10月20日と21日に送付された2通の書簡をブルームバーグ法律事務所が精査した。
フォ…
Loading...