How to protect yourself from bogus AI court filings by your attorney

by Donald ‘C-Note’ Hooker 

As a retired jailhouse lawyer, I will continue to sound the alarm when I read articles about court filings that are bogus as the result of using AI. My concerns come from several motivations. As a person who believes in capitalism, I wonder why are we turning over our money to an overly paid specialized profession when I could produce the same legal product for pennies on the dollar? 

Unlike other branches of the legal profession, criminal law deals with a whole lot of people who cannot afford such services – in other words, poor people. Now, that’s a crime. In the realm of criminal law, poor criminal defendants are likely to be caught up in a farm mill of overly stressed public defenders with too many clients for one attorney to handle. AI, which can spit out a court document in seconds, solves a whole lot of problems that good ol’ cumbersome, manual, human research cannot. That’s to say, it saves time. But what good is saving time, when AI has consistently shown not to be up to the task when it comes to legalese? 

On Oct. 14, 2025, Fox News Top Stories reported, “Federal judge fines, reprimands lawyer who used AI to draft court filings,” by Stephen Sorace. “U.S. District Judge Terry Moorer penned an 18-page order that chastised the conduct of Loxley-based lawyer James Johnson,” wrote Sorace. Judge Moorer was beyond critical; he noted the conduct of this attorney was just wrong: “Mr. Johnson’s misconduct was more than mere recklessness. … The insertion of bogus citations is not a mere typographical error, nor the subject of reasonable debate; it is just wrong.” 

In an article I wrote earlier this year, “How prisoners aren’t educated on the failures of AI in court proceedings,” I map out the pervasiveness of this problem, most notably by highly paid, lifetime appointed federal judges. U.S. District Judge Moorer equally felt an extreme need to state out loud that this is not only an epidemic, but a five alarm problem: “The improper use of generative AI is a problem that sadly is not going away despite the general knowledge in the legal community that AI can hallucinate and make up cases. … Somehow the message still has not been hammered home, as the epidemic of citing fake cases continues unabated,” wrote Judge Moorer. 

Liberty is the greatest aspect of the American experiment. It is from liberty that the Bill of Rights are derived, such as freedom of religion, freedom of speech and so on. It’s all about having the liberty to express oneself, in thought and in belief systems. While these are ethereal freedoms, the loss of physical freedom due to constraints is not something of the mind, but of the body. The loss of one’s physical liberty has no legal comparison. As a Black American incarcerated for the past 29 years for a crime I didn’t commit, I understand this fully. 

Statistics show that Black incarceration rates are six times that of whites and two times that of Hispanics. But how is that, when both of these groups outnumber the Black population in America? So this is personal for me. And it is personal for anybody who is caught up in the criminal justice system. But to see countless Black Americans in the meat grinder of city and county jails fighting criminal cases is quite disheartening, especially knowing these defendants are generally poor and cannot adequately defend themselves in court. 

Now comes generative AI, which seemingly poses a solution to a long held problem in the criminal justice system – too many defendants for too few attorneys. Unfortunately, the miracle of generative AI is not ready for primetime in the legal profession, and somebody needs to tell this to Sam Altman of OpenAI. 

One solution is to create a generative AI solely for the legal profession. Such a large language model must be trained solely on constitutional, statutory and decisional law. If someone is reading this, it does not come from a place of people, including myself, who are using free versions of AI. We are paying, premium users, and the results generative AI is producing are bogus – made up case law or case law whose outcomes we are alleging in our legal filings are not true. 

From my understanding, they already have such apps; however, they are producing the same AI hallucinated results as ChatGPT. In “How prisoners aren’t educated on the failures of AI in court proceedings,” I noted, “The American Bar Association holds the position, ‘For bar-admitted attorneys, these erroneous court submissions are not taken lightly. Lawyers are responsible for the veracity of all information included in court filings, including if it includes AI-generated materials.’” I also noted that one judge wrote, “No reasonably competent attorney should outsource research and writing to this technology without any attempt to verify the accuracy of that material.” 

For criminal defendants represented by overworked public defenders, it will be difficult to fact check their court filings. However, for those already convicted, it is much easier to fact check appeal court legal briefs by your appellate attorney. In the legal case that spurred me to write this article, Sorace cited in his article, “Judge Moorer granted Johnson’s client permission to represent himself following the AI misconduct.” 

C-Note is known as the world’s most prolific prison artist, who has written for the John Jay College of Criminal Justice, Mprisond Thotz. Inmate Blogger and more. To stay updated on C-Note’s parole board journey or Mumia Abu-Jamal’s reporting on prison elder abuse, visit the website P.E.A.C.E. (Parole Elder Abuse Concerns Everyone), www.paroleelderabuse.org. He’s also available for Zoom and other forms of call-ins at meetings or events. Message him on the GTL Getting Out app: Donald Hooker, K94063 (A2-150), P.O. Box 4430, Lancaster, CA 93539. 

The post How to protect yourself from bogus AI court filings by your attorney appeared first on San Francisco Bay View.