Pro Se Advocacy in the AI Era: Benefits, Challenges, and Ethical Implications
2.10.2026
The recent advancements in artificial intelligence are attracting many to use the technology in everyday practice. Such technology presents new opportunities and challenges for people without legal representation and the legal system. That shift reflects potential for improving accessibility to the legal system, and it is reshaping the dynamics of legal advocacy. Despite such advantages, these developments come with notable challenges such as accuracy, reliability, and the unauthorized practice of law.
The majority of pro se litigants are pro se plaintiffs, who consisted of 28% of all cases in the federal courts between 1999 and 2018.[1] A 2006 survey revealed that there were 3,303 pro se litigants in New York City Family Court and New York City Housing Court.[2] It was estimated that the number of pro se litigants in the New York State Unified Court System is around 83,000 as of December 2024.[3] These numbers indicate that pro se litigants are common across New York State and that it is an increasing phenomenon.
A recent survey by the National Center for State Courts indicated that several barriers face pro se litigants in navigating e-filing systems in civil courts.[4] Some of these barriers are the lack of understanding of their legal rights as well as lack of access to technology.[5] Other barriers also include the rising costs of litigation coupled with decreased funding for legal services, and the rise of do-it-yourself legal services.[6]
Pro se litigants are usually considered to be unduly burdensome on judges, clerks, and court processes, as they tend to take more time and require more assistance to comply with legal formalities and requirements.[7] As pro se litigants seem to burden the court systems, technology could be used to help overwhelmed courts. AI technologies like ChatGPT serve as supplemental tools to assist with tasks such as drafting legal letters to courts preparing defenses (e.g., parking tickets), and addressing procedural requirements, such as opening statements.[8]
Sateesh Nori, executive director of Justfix, a housing justice technology nonprofit, who also runs an eviction defense and tenant protection clinic at the New York University School of Law, sees potential in AI technology, “particularly in legal processes that are built on forms and letters. … ‘This is like the invention of fire,’” he said.[9]
For individuals who often face barriers to obtaining affordable legal representation, AI tools are an appealing alternative. At a legal clinic in San Francisco, one client was able to use AI to read and summarize hundreds of pages of documents. The attorney at the clinic was then able to direct him on next steps, all in under a half hour.[10]
However, the demand for resources for pro se litigants highlights the critical need to make legal representation more accessible and affordable to self-represented individuals. Despite utilizing AI technologies, pro se litigants need assistance in employing these tools correctly. The AI tools themselves have warnings prohibiting their use for the unauthorized practice of law or legal advice without confirmation from a qualified person.[11] Therefore, pro se litigants need to engage in extensive fact-checking and cross-referencing with other resources to ensure reliability. In one case in Missouri, a pro se litigant was fined $10,000 after using AI to generate fake citations in briefs.[12] This suggests that while AI tools like ChatGPT can be a helpful starting point, it is not yet equipped to replace thorough legal research or experienced counsel.
The unreliability of AI combined with its ready accessibility poses new challenges for the court system. This is illustrated, in Dewald v. MassMutual Metro, where the Appellate Division, First Department allowed a pro se appellate to submit a pre-recorded oral argument.[13] To the appellate court’s surprise, the pro se appellant admitted using artificial intelligence to generate an avatar.[14] Prior to the submission to the Appellate Division, the pro se appellant did not disclose the use of artificial intelligence to generate an avatar that was reading a scripted argument, which may have also been AI generated.
In order to overcome these challenges, several initiatives employing AI are attempting to democratize access to legal support. The Northern District of New York, for example, has launched Pro Se Pal, a chatbot to answer questions, guide users though procedure, and help with court paperwork.[15] Pro Se Pal cannot give legal advice, but it can assist in the preparation and drafting of papers for defense of a civil action in the Northern District of New York.[16]
Legal aid attorneys, clinics, and institutions such as New York University have been experimenting with AI tools as hotlines to respond quickly to thousands of legal questions from pro se litigants.[17] Another initiative, Housing Court Answers, in collaboration with New York University, developed an AI assistant to help New York City tenants to understand and exercise their rights.[18]
While efficiency and accessibility are some benefits of using such technology, AI still poses the danger of hallucination, which can create fake citations and improper legal advice. Courts across the country remain increasingly attentive to the risk of AI-generated filing and remain understandably cautious about the use of AI technologies by pro se litigants. Several federal judges nationwide have ordered pro se litigants to disclose the use of AI in preparing a case as well as certification of citation accuracy.[19] While sanctions and rules for certifications of AI-generated content exist, these have not stopped attorneys or pro se litigants from using AI tools in preparing legal documents. Some courts have banned the use of generative AI altogether by both attorneys and pro se parties.[20] While many courts have now adopted certification by litigants if any portion of their pleadings used AI technology, it remains difficult for courts to verify such use. The lack of adequate and reliable AI detection poses a risk, especially in small claim courts, where many litigants represent themselves.
In a recent case in Kings County, the court addressed a pro se defendant’s use of AI generated motions, which then meant the court and plaintiff’s counsel had to take time to verify the citations.[21] The court noted that attorneys as well as pro se litigants are expected to adhere to the rules regarding the use of artificial intelligence.[22] In another Kings County case, the court stated that “a self-represented litigant acquires no greater rights than any other litigant.”[23]
In a 2024 report, NYSBA’s Task Force on Artificial Intelligence acknowledged that AI could help underserved populations access legal guidance, but its use could also cross the line into the unauthorized practice of law. The task force also noted concerns that “individuals in underserved communities or with limited financial means will be relegated to inferior AI-powered technology.”[24] The task force posed the question, “Are the people, who otherwise would not have legal counsel, better served by at least having a chatbot to assist them?”[25]
Eventually, a new set of rules governing submissions of AI-generated pleadings in legal proceedings will need to be established. Pro se affirmation, similar to the Safe Harbor Rule,11 should accompany any pleadings submitted by the pro se litigants to confirm and disclose whether any part of the pleading was generated by artificial intelligence. Such affirmation is particularly important when it comes to citations generated by AI language models. Ultimately, the courts must determine whether such reliance on AI by pro se litigants mounts to the unauthorized practice of law, and not merely a guidance tool available for the public.
For pro se litigants, specifically, the court system’s adoption of citation checking tools such as Westlaw Edge or Lexis+ AI may help instill more confidence for pro se litigants in their filings and also lower the risk of courts creating precedent using nonexistent cases. Such adoption would require significant resources and training for pro se litigants. Until these resources exist, courts and legal services may need to increase their efforts in warning self-represented litigants about the risks of using free AI platforms, and in pointing them in the direction of trusted sources or legal databases.
Overall, AI technology such as ChatGPT remain in its infancy. While there are many technological and ethical challenges, AI tools give hope to attorneys and pro se litigants to increase efficiency and improve access to justice.
Mostafa Soliman is a counsel at Coffey Modica LLP. He served as a legal fellow with Equal Justice Works AmeriCorps in Western New York.
Endnotes:
[1] Mark D. Gough and Emily S. Taylor Poppe, (Un)Changing Rates of Pro Se Litigation in Federal Court. 45 Law & Social Inquiry 3 (2020) doi:10.1017/lsi.2019.69.
[2] The New York State Unified Court System’s Office of the Deputy Chief Administrative Judge for Justice Initiatives, Self-Represented Litigants in the New York City Family Court and New York City Housing Court (December 2005), https://ww2.nycourts.gov/sites/default/files/document/files/2018-06/AJJI_SelfRep06.pdf.
[3] New York Unified Court System, Electronic Filing in the New York State Courts, Report of the Chief Administrative Judge to the Legislature, the Governor, and the Chief Judge of the State of New York 2025, https://iappscontent.courts.state.ny.us/NYSCEF/live/help/EFileReport.pdf?utm_source=chatgpt.com.
[4] National Center for State Courts, Self-Represented Efiling: Surveying the Accessible Implementations, May 2022, https://ncsc.contentdm.oclc.org/digital/api/collection/accessfair/id/966/download.
[5] Id.
[6] Kelsey Whitt, The Split on Sanctioning Pro Se Litigants Under 28 U.S.C. 1927: Choose Wisely When Picking a Side, Eighth Circuit, 73 Mo. L. Rev. (2008).
[7] Van Wormer and Nina Ingwer, Help at Your Fingertips: A Twenty-First Century Response to the Pro Se.
Phenomenon, 60 Vanderbilt Law Review 3, at 983 (2007), https://scholarship.law.vanderbilt.edu/vlr/vol60/iss3/6/.
[8] Mostafa Soliman, Two Sides of the AI Coin: Pro se Litigants and Attorneys on ChatGPT A Survey-based Study (April 26, 2025). available at SSRN: https://ssrn.com/abstract=5231375 or http://dx.doi.org/10.2139/ssrn.5231375 for full access to survey data. The survey is based on responses received from pro se litigants who agreed to share their experience using ChatGPT in actual legal proceedings. Only seven individuals were willing to share such information. Due to the limited responses, this study gives a narrow prospective of pro litigants’ behavior in using AI technology in legal proceedings.
[9] Marco Poggio, Gen AI Shows Promise – And Peril – For Pro Se Litigants, Law360 (May 3, 2024), https://www.law360.com/articles/1812918/gen-ai-shows-promise-and-peril-for-pro-se-litigants.
[10] Id.
[11] Id.
[12] Id.
[13] Appellate Division, First Department, Live Stream, (YouTube, March 26, 2025), https://www.youtube.com/watch?v=Ctv4ZQRZgbA&t=1170s.
[14] Id.
[15] Northern District of New York Federal Court Bar Association, Pro Se Assistance Program, https://ndnyfcba.org/about-us/pro-se-assistance-program.
[16] Id.
[17] Poggio, supra note 8.
[18] Housing Court Answers Launches First AI Assistant for NYC Tenants in Collaboration With NYU and Josef, Josef (Jan. 14, 2025), https://joseflegal.com/blog/housing-court-answers-launches-first-ai-assistant-for-nyc-tenants-in-collaboration-with-nyu-and-josef/.
[19] Sarah Martinson, Judges’ AI Orders Keep Trickling in as Fake Citations Persist, Portfolio Media, Law 360 (July 21, 2025), https://www.crowell.com/a/web/hATTWfj2h9N1Tf42uN3QFg/judges-ai-orders-keep-trickling-in-as-fake-citations-persist.pdf.
[20] Poggio, supra note 9.
[21] Augustin v. Formula 3 Brooklyn Inc., 2025 N.Y. Slip Op. 51113(U).
[22] Id.
[23] Kolomenskava v. Meta Platforms, Inc., 430/2025, citing Duffen v. State, 245 AD2d 653, 665 N.Y.S.2d 978; Brooks v. Inn at Saratoga Assn., 188 AD2d 921, 591 N.Y.S.2d 625 (3rd Dep’t 1992).
[24] N.Y. St. Bar Ass’n, Report and Recommendations of the New York State Bar Association Task Force on Artificial Intelligence, NYSBA (April 2025), p. 24, https://nysba.org/wp-content/uploads/2022/03/2024-April-Report-and-Recommendations-of-the-Task-Force-on-Artificial-Intelligence.pdf. The report includes guidelines for the use of AI and Generative AI in legal practice.
[25] Id. at 44.
