SCRANTON – As the use of artificial intelligence (AI) proliferates, one federal judge in Pennsylvania has made it clear that use of the tool in her courtroom will only occur if it is disclosed and what portions of legal documents it was used to generate are clearly specified.
U.S. District Court for the Middle District of Pennsylvania Judge Karoline Mehalchick issued a civil practice order to that effect on Sept. 9, outlining her rationale and the stipulations by which counsel need to abide if they plan to use AI for documents submitted in her court.
“Increased use of Artificial Intelligence (“AI”), particularly Generative AI (including, but not limited to, OpenAI’s ChatGPT or Google’s Bard), in the practice of law raises a number of practical concerns for the Court, including the risk that the generative AI tool might generate legally or factually incorrect information, or that it might create unsupported or nonexistent legal citations. As such, any party, whether appearing pro se or through counsel, who utilizes any generative AI tool in the preparation of any document to be filed in any matter pending before Judge Mehalchick, must include with the document a Certificate of Use of Generative AI in which the party must disclose and certify: 1)The specific AI tool that was used; 2) The portions of the filing prepared by the AI program; and 3) That a person has checked the accuracy of any portion of the document generated by AI, including all citations and legal authority,” Mehalchick’s order read.
“Failure to comply with this Order may result in sanctions. Further, all parties and counsel are directed to review the conclusions on Pages 15 and 16 of the Joint Formal Opinion of the Pennsylvania Bar Association and Philadelphia Bar Association regarding the use of Artificial Intelligence and be mindful of their ethical and professional obligations before this Court.”
In that opinion, it makes clear that when using AI, a lawyer “must ensure that any client information and materials remain confidential and safeguard that information to ensure that it is protected from breaches, data loss and other risks.”
The opinion further advises that attorneys be truthful and accurate, verify all citations and the accuracy of cited materials, be proficient in using AI technology, maintain confidentiality, identify conflicts of interest, communicate with their clients, be assured that their AI-generated material is accurate and unbiased, adhere to ethical standards, utilize proper judgment and billing practices and maintain transparency.
“Artificial intelligence and generative AI tools, like any tool in a lawyer’s arsenal, must be used with knowledge of their potential and an awareness of the risks and benefits the technology offers. They are to be used cautiously and in conjunction with a lawyer’s careful review of the ‘work product’ that those types of tools create. These tools do not replace personal reviews of cases, statutes and other legislative materials. Additionally, although AI may offer increased productivity, it must be accomplished by utilizing tools to protect and safeguard confidential client information,” the opinion stated.
Attorneys elsewhere who ran afoul of improperly using AI have done so with costly results.
In a New York federal appellate court earlier this year, an attorney was faced with the prospect of paying more than $8,000 in sanctions, after using ChatGPT to find and cite a case to support her client’s claims – only for the court to later learn that the case itself did not exist.
The Pennsylvania Record inquired to Mehalchick’s chambers regarding her order on the use of AI in her courtroom, but a spokesperson for her office explained that judges usually do not comment to the media on such orders and likewise, neither the judge nor any personnel from her chambers would be able to comment.
From the Pennsylvania Record: Reach Courts Reporter Nicholas Malfitano at nick.malfitano@therecordinc.com