Introduction: In February 2022, Allbaugh, a pro se litigant (and former attorney) sued the University of Scranton after not being selected for a position, alleging gender discrimination. Throughout court proceedings, an unexpected issue arose - the misuse of generative Artificial Intelligence (AI) in court filings.
Key Facts:
After Allbaugh filed the complaint, the University moved to dismiss for failure to state a claim. The Judge recommended granting the University of Scranton's motion under the Twombly-Iqbal standard while allowing Allbaugh leave to amend.
In his amended complaint and objections, Allbaugh included a potentially fictitious citation generated by AI prompting the Court to issue a show-cause.
Allbaugh responded by admitting to using AI and the Court found him in violation of several authorities:
The Judge’s standing order regarding generative AI {1} (even though it was issued after Allbaugh's initial claim);
Rule 11 of the Federal Rules of Civil Procedure {2}, which requires parties to certify that their filings are not submitting false or fictitious material;
The Pennsylvania Rules of Professional Conduct {3}, which applies to attorneys.
Rulings: The Court OVERRULED Allbaugh's objections, GRANTED the University’s motion to dismiss, and SANCTIONED Allbaugh, ordering him to pay a $1,000 penalty for violating the Court’s standards
This case illustrates the growing dilemma between technological innovation and ethical responsibility within the justice system. While AI tools have the potential to make legal processes more accessible - especially for pro se litigants - they can also magnify risks when used without proper understanding, verification and disclosure. For individuals seeking justice without legal representation, AI can either be an empowering tools or a liability.
What do you think?
Is there a way for courts to “ethically” embrace AI to support pro se litigants while preventing misuse?
What safeguards are necessary to maintain the integrity of the judicial process as technology continues to evolve?
Footnotes:
{1} Judge Mehalchick Standing Order Regarding Use of AI: https://www.pamd.uscourts.gov/content/judge-karoline-mehalchick
{2} Federal Rules of Civil Procedure (Rule 11): https://www.law.cornell.edu/rules/frcp/rule_11
{3} Pennsylvania Rules of Professional Conduct: https://www.padisciplinaryboard.org/for-attorneys/rules/rule/3/the-rules-of-professional-conduct
I don't think the burden should be on the court to support pro se litigants. If someone is deciding to represent themselves, I think that burden should be on them to do their research before appearing in court, including knowing the dangers of using AI as a source. I also think having any sort of AI support option would open the court up to a lot of different issues. Especially when we look at the case in your post. If someone were to use AI that was supported by the court and then later discover the cited cases were "hallucinated," I think you could argue that the court could be partially responsible. We talked about the use of AI a little bit in our legal research class and our professor talked about a similar case in the Utah Court of Appeals where two lawyers had used ChatGPT for a brief and that brief included hallucinated case law. Our professor pointed out the importance of not relying on AI for facts, and to always double check any citations it gives. I think that rule can be applied as technology continues to evolve. Don't assume it's always right. Even now when you use search engines, you need to be sure you phrase your questions correctly so you aren't searching with a confirmation bias. Learning how to use technology correctly is a useful skill and something we always should be willing to improve on.
ReplyDeleteI agree with McKenna completely; AI is not a tool to be used lightly, and it is not reasonable to expect the court to support or grant leniency to pro se litigants (or other legal counsel) who do not utilize AI correctly. Obvious AI errors in a party’s filings that get accepted by the court require corrective measures to be taken later when discovered, opens up the likely possibility of appeals, and overall extends the time taken to resolve a dispute. AI that is incorrectly used in the litigation process wastes time, money, and resources.
DeleteAI can be an effective tool, but there is no excuse or justification for using it to cut corners. If one is to use AI, they need to be fully in control of the research and decision-making process. Pro se litigants have a responsibility to abide by the best practices available for AI usage, and the court has a responsibility to hold all parties equally accountable.
In my current job I have witnessed individuals representing themselves using AI to help them in making arguments for their cases, and can see how this can not always work. An important safeguard I think may be helpful is restricting where and how AI can be used within the legal system. While AI can be a valuable tool, particularly for pro se litigants who may lack access to legal resources, its use should be carefully limited. For example, AI could be permitted to help individuals understand the procedural aspects of a case, such as filing requirements or deadlines, but not to conduct legal research or draft filings. This limitation would help prevent the kind of misuse or inaccuracies like the one presented in your case. I think that it is important that as technology evolves so should our education on the use of it.
ReplyDeleteHi Erica, I think you make an excellent point about the need for safeguards around AI use in the legal system.
DeleteYour idea of limiting AI to procedural assistance rather than legal drafting is a good approach. I think this solution offers empowerment and protection for pro se litigants because understanding how to navigate the legal process like deadlines, forms, service requirements, etc. is half the battle.
To address Marina's question, this raises questions about where to draw the line. One suggestion to help courts "ethically" embrace AI is to demand that AI tools disclose their limitations, cite their sources, and guide users toward verified information rather than “hallucinated” legal claims.
I agree with some other comments that it shouldn't be the court's responsibility to support pro se litigants or to endorse AI use, given the high risk of errors. As others have mentioned, this could lead to many issues for the court. With all the technology targeting attorneys to use and purchase AI tools, it seems necessary to implement safeguards to protect the integrity of the judicial process. Clear rules and laws should be established that restrict where AI can be utilized. This will also protect the public from AI tools that claim to offer legal advice or legal help. Serious Penalties and sanctions should be employed by the court to deter individuals in the future and to set an example to others of the seriousness of the error.
ReplyDeleteI'm actually on the fence a bit about this issue. I'm not super Pro-AI, but sometimes I think it's useful for creating outlines for formal writing. In this case however, from a legal standpoint, I can see how this may be a bit controversial. The fault immediately goes on whoever is choosing to use AI, however, if the information they choose to copy is incorrect. The courts should leave the liability on the people, I don't agree with restrictions on this. At the end of the day as technology advances, it's the job of both judges/jury and the people to educate themselves on how to use AI and detect when it's not accurate.
ReplyDelete