By Bill Hartzer · Published · How this site is researched
The six topic pages of this record are organized by who used the system: an expert, a lawyer, a judge, a party, a rulemaker. Readers rarely arrive with that question. They arrive asking whether an expert's prompts can be ordered produced, what a court is likely to do with a report that cites a source no one can find, which courts now require a certificate about AI use, or what it takes to challenge a video as a deepfake. The guides below answer those questions from the matters in the record.
Each guide follows the record's rules. It states what particular tribunals did, names and links every matter it relies on, and notes where a matter is pending, unpublished or interlocutory. It does not characterize a holding as a rule of law, and it is not legal advice: the law differs between courts and changes, and a reader with a live matter should read the primary documents the entries link and take advice in the jurisdiction concerned. Where the matters point in different directions, the guide says so rather than choosing between them.
Most of the United States matters turn on a small number of rules. Discovery of an expert's materials runs through Federal Rule of Civil Procedure 26, which requires disclosure of the facts or data an expert considered and protects drafts of the report. Challenges to the reliability of an opinion run through Federal Rule of Evidence 702. Filings with fabricated authority are usually addressed under Rule 11 or a court's inherent power. State courts and other legal systems have their own versions of each, and the guides say which system a matter arose in.
The guides are kept with the record. When an entry is updated, the guides that discuss it are re-read against the change, and each guide carries the date it was last checked. The clauses of the Standard that each guide engages are linked to their own pages, which list every matter in the record that engaged the same clause.
AI Disclosure and Certification Rules, Compared
No single rule governs. The instruments in this record fall into two families. Some require a filer to disclose that a generative system was used, as the Cherokee County standing order, California SB 574 and the draft Indian regulations do. Others require only that whatever a system produced be verified before filing, as Connecticut's Practice Book Section 4-9 and the Tenth Circuit's proposal do. Of the instruments in force, only the New South Wales practice note reaches expert reports by name.
Deepfake Claims About Evidence: How Tribunals Have Ruled
Tribunals have treated the two directions differently. Where evidence was shown to be fabricated, as in Mendones v. Cushman & Wakefield and Adams v. Anderson, the consequences were dismissal with prejudice and a contempt sentence. Where a party only asserted that genuine evidence might be fake, as in State v. Amyda and Burnley v. Valentin, tribunals asked for an evidentiary predicate and, finding none, admitted the evidence. A proponent who cannot authenticate a recording can still lose it.
Discovery of AI Prompts and Chat Logs: What the Record Shows
The answer has depended on who did the prompting and why. In Villanueva v. Las Vegas Metropolitan Police Department, a federal district judge ordered an expert's ChatGPT case summaries produced as facts or data considered but protected his prompts and draft output. In Conservation Law Foundation v. Shell, a magistrate judge ordered an expert team's prompts produced as methodology; that order is stayed. For parties, United States v. Heppner rejected protection, while several trial courts protected party use as work product. No appellate court has resolved the question.
Expert Reports That Cite Sources That Do Not Exist
The outcomes in this record range from a report struck before trial to an opinion left entirely in place. Tribunals have excluded declarations, withdrawn reports in mid-trial, set a court-appointed witness's fee at zero, found an opinion not credible, reduced its weight, and in several matters left the defect to cross-examination. The result has tracked the kind of defect, how central the material was, and whether the author could account for it, rather than the mere use of a system.
Related reference
Terms used across the guides are defined in the glossary. The record itself is indexed on the AI in the Courtroom page, by topic and by date, and the obligations the guides refer to are set out in the Standard, one page per clause.