AI and the Legal Profession
AI and the Courts[edit | edit source]
Judicial Attitudes Toward AI[edit | edit source]
Most U.S. Judges Have Never Tried AI Tools[edit | edit source]
A recent survey of U.S. judges by the U.S. National Judicial College (@thenjc) showed that three quarters of U.S. judges have never tried AI tools like ChatGPT, but that almost one fifth had tired it and found it helpful:
Of the 332 judges who responded, nearly 76 percent said they hadn’t tried it. About 7 percent said they had tried it and don’t like it. About 17 percent said had tried it and do like it.
Ed Cohen (linkedin profile), "Most judges haven’t tried ChatGPT, and they aren’t impressed" (July 21, 2023) At least one judge expressed grave concerns:
“At the rate we’re going,” the judge wrote, “‘chatbots’ will replace judges and decisions will be made solely on algorithms. When people’s lives hang in the balance, it is beyond scary to think that the human factor may be removed entirely!” Limiting AI to merely “research and other legal tasks,” as mentioned in the poll question, “is only the beginning to a disastrous end,” the judge predicted.
Thanks to Damien Riehl (damienriehl@law.builders, @damienriehl) for the pointer.
Stories of Judges Using AI Tools[edit | edit source]
- Judge discards ChatGPT "research memo" that included made up cites: A bankruptcy judge tried using ChatGPT for background research on social media marketing (an issue relevant to the dispute) but dumped the research after finding that ChatGPT "listed five sources in all" but "none of the five seem to exist." The case is In re: Vital Pharmaceutical in the Southern District of Florida Bankruptcy Court, Case no. 22-17842-PDR. Thanks to Eugene Volokh of the Volokh Conspiracy.
Our Take[edit | edit source]
For our coverage of how judges are reacting to AI see
Are Courts 🧑⚖️ Unfairly Blocking AI 🤖?
Cases where attorneys have filed AI-generated documents[edit | edit source]
| Case name | Jurisdiction | Issue | Result | Docket |
|---|---|---|---|---|
| Mata v. Avianca, Inc. | U.S. SDNY | brief with multiple cites to "fabricated" cases
copies of fabricated cases |
sanctions against attorneys in the amount of $5000.00 | https://www.courtlistener.com/docket/63107798/mata-v-avianca-inc/ |
| Al-Qarqani v. Chevron Corporation | U.S. 9th Cir. | fabricated news story | sanctions against attorney for attorneys fees in the amount of $250,000 | Court of Appeals for the Ninth Circuit: Al-Qarqani v. Chevron Corporation (19-17074)
https://www.courtlistener.com/docket/64927677/al-qarqani-v-chevron-corporation/ |
| ?? | U.K. Manchester | Media reports that A "litigant in person" (in the U.S. and Canada, a "self represented" or "pro se" party) submitted made up case names in a civil case | "The judge accepted the misleading submissions were inadvertent and did not penalise the litigant"
LiP presents false citations to court after asking ChatGPT, Law Society Gazette, 2023-05-28 |
|
| Gates v. Zalaya Chavez | District Court, El Paso County, Colorado | Motion with citations to fabricated cases. | Stipulation to discipline and suspension for one year and one day, with 90 days to be served and the remainder to be stayed upon the attorney's successful completion of a 2-year period of probation.
|
|
| US v. Michael Cohen | U.S. SDNY | Motion with three citations that referred to nonexistent cases. | Order to Show Cause: https://storage.courtlistener.com/recap/gov.uscourts.nysd.499666/gov.uscourts.nysd.499666.96.0.pdf
|
United States v. Cohen (1:18-cr-00602)https://www.courtlistener.com/docket/7709756/united-states-v-cohen/ |
| Park v. Kim | U.S. Court of Appeals for the 2nd Circuit | Reply brief with citation to nonexistent court opinion. | The attorney was referred to the Court's Grievance Panel, and further ordered to furnish a copy of the appellate court's decision to her client.
Order: https://www.courtlistener.com/docket/66919280/178/1/park-v-kim/
|
Appellate Case: Case 22-2057
https://www.courtlistener.com/docket/66919280/park-v-kim/
|
| Kruse v. Karlen | Missouri Court of Appeals | Pro se appellant submitted an appellate brief with fictitious citations. | Court required appellant to pay $10,000 in "damages for filing a frivolous appeal."
|
Appellate Case: ED111172 - Appeal from the Circuit Court of St. Charles County |
Court orders regarding filing AI-generated materials[edit | edit source]
| Country | Court | Judge | Date | summary | link |
|---|---|---|---|---|---|
| USA | TX ND | Starr | 2023-05-30 | any language drafted by generative artificial intelligence will be checked for accuracy, using print reporters or traditional legal databases, by a human being | Mandatory Certification Regarding Generative Artificial Intelligence |
| USA | IL ND | Fuentes | 2023-06-08 | Any party using any generative AI tool to conduct legal research or to draft documents for filing with the Court must disclose in the filing that AI was used, with the disclosure including the specific AI tool and the manner in which it was used | Standing Order for Civil Cases Before Magistrate Judge Fuentes |
| Canada | Manitoba King's Bench | 2023-06-23 | "when artificial intelligence has been used in the preparation of materials filed with the court, the materials must indicate how artificial intelligence was used" | ||
| USA | Court of Int'l Trade | Vaden | 2023-06-08 | "any submission in a case assigned to Judge Vaden that contains text drafted with the assistance of a generative artificial intelligence program on the basis of natural language prompts, including but not limited to ChatGPT and Google Bard, must be accompanied by:
(1) A disclosure notice that identifies the program used and the specific portions of text that have been so drafted; (2) A certification that the use of such program has not resulted in the disclosure of any confidential or business proprietary information to any unauthorized party; and it is further " |
Order on Artificial Intelligence |
| USA | MT D | Molloy | 2023-06-22 | Note: by its terms, this order applies only to a specific case
"Use of artificial intelligence automated drafting programs, such as Chat GPT, is prohibited" |
Order Granting Admission Pro Hac Vice |
| USA | PA ED | Baylson | 2023-06-06 | If any attorney for a party, or a pro se party, has used Artificial Intelligence (“AI”) in the preparation of any complaint, answer, motion, brief, or other paper, filed with the Court, and assigned to Judge Michael M. Baylson, MUST, in a clear and plain factual statement, disclose that AI has been used in any way in the preparation of the filing, and CERTIFY, that each and every citation to the law or the record in the paper, has been verified as accurate. | Standing Order Re: Artificial Intelligence ("AI") In Cases Assigned to Judge Baylson |
| Canada | Yukon Supreme Court | 2023-06-26 | "if any counsel or party relies on artificial intelligence (such as ChatGPT or any other artificial intelligence platform) for their legal research or submissions in any matter and in any form before the Court, they must advise the Court of the tool used and for what purpose." | Use of Artificial Intelligence Tools | |
| USA | TX 394th D | Ferguson | 2023-06-09 | (1) "all language, quotations, sources, citations, arguments, and legal analysis created or contributed to by generative artificial intelligence were before submission verified as accurate through traditional (non-AI) legal sources by an attorney licensed to practice law in the State of Texas"
(2) anyone submitting any materials to court "understands and acknowledges that they are and will be held responsible and potentially sanctioned for their or their co-counsel’s failure to comply with this Order." |
Standing Order Regarding Use of Artificial Intelligence |
| USA | TX ND Bk | 2023-06-21 | If any portion of a pleading or other paper filed on the Court’s docket has been drafted utilizing generative artificial intelligence, including but not limited to ChatGPT, Harvey.AI, or Google Bard, the Court requires that all attorneys and pro se litigants filing such pleadings or other papers verify that any language that was generated was checked for accuracy, using print reporters, traditional legal databases, or other reliable means. | General Order 2023-03 In Re: Pleadings Using Generative Artificial Intelligence | |
| USA | NY SD | Subramanian | 2023-07-29 | Counsel is responsible for providing the Court with complete and accurate representations of the record, the procedural history of the case, and any cited legal authorities. Use of ChatGPT or other such tools is not prohibited, but counsel must at all times personally confirm for themselves the accuracy of any research conducted by these means. At all times, counsel—and specifically designated Lead Trial Counsel—bears responsibility for any filings made by the party that counsel represents. | Individual Practices in Civil Cases |
| Canada | Alberta Courts | 2023-10-06 | In the interest of maintaining the highest standards of accuracy and authenticity, any AI-generated submissions must be verified with meaningful human control. Verification can be achieved through cross-referencing with reliable legal databases, ensuring that the citations and their content hold up to scrutiny. | Ensuring the Integrity of Court Submissions when Using Large Language Models | |
| USA | US District Court - Hawaii | Watson, Seabright, Kobayashi, & Otake | 2023-11-14 | If any counsel or prose party submits to the court any filing or submission generated by an unverified source, that attorney or pro se party must submit a declaration concurrently with that material captioned "Reliance on Unverified Source" that: (1) advises the court that counsel or the pro se party has relied on one or more unverified sources; and (2) verifies that the counsel or pro se party has confirmed that any such material is not fictitious. The scope of the required declaration is that required by Rule 11 of the Federal Rules of Civil Procedure. This order does not affect the use of basic research tools such as Westlaw, Lexis, or Bloomberg, and no declaration is required if all sources can be located on such well-accepted basic research tools. | General Order 23-1 |
For updates, please subscribe to our substack newsletter.
Thanks to Carolyn Elefant (@carolynelefant), Stephen Embry, Colin Lachance (@colinlachance), Lori Shemka (@shemkalegal) Amy Salyzyn (@AmySalyzyn) (who am I missing?)
Reactions to Courts barring use of Generative AI[edit | edit source]
Summary of various positions:
| Restricting generative AI for court filings is moot: | "Kenneth J. Withers, deputy executive director of the Sedona Conference pointed out that the question of whether generative AI played a part is moot, “because the brief is inept whether or not it was written by an associate by a partner or by generative AI” and should be considered so regardless of whether a human or a machine wrote it. | ||
| Restrictions on generative AI are overbroad, and sweep up, e.g. spell checkers | "Does this mean that if you’re using a program like Grammarly or if you’re using a translation program, which no one really thought was objectionable before do you suddenly have to certify that this has been done?” | ||
| Requiring disclosure could violate Attorney Work Product protection | judicial orders requiring disclosures of generative AI use, and the how review and certification of research is done, may risk the violating work-product privilege, which permits an attorney to withhold the strategies and materials prepared in anticipation of litigation from production |
4 Generative AI Issues That Are Likely Keeping Judges Up at Night, Isha Marathe (@IMarathe) 2023-08-10
Stephen Embry Judicial Treatment of ChatGPT: Throwing the Baby Out With the Bath?
Effect of AI on Legal Work[edit | edit source]
Recent analysis argues that the Legal Industry is at the top of the list of industries "most exposed" to Artificial Intelligence
How will Language Modelers like ChatGPT Affect Occupations and Industries?
The Potentially Large Effects of Artificial Intelligence on Economic Growth