The Right to Privacy

There are many opportunities for us to learn from famous people. Their high profiles draw attention and make them examples in the news. Today, with the vast array of social media, it is more apparent than ever. I am not one to pick on celebrities, and there is way too much willingness to punch folks when they are down.

This brings me to Tiger Woods, a Floridian of some fame. In March 2026, Golf Digest made a blunt assessment: "Something is wrong with Tiger Woods." It concluded that this was neither subtle nor novel, but concluded, instead, that it was "sad and disconcerting."

That coverage followed a "two-vehicle crash" after which Mr. Woods was arrested for driving while impaired. Golf Digest noted his history of injuries, chronic pain, and "the pharmaceutical architecture required just to get through a day." As I read that, it occurred to me that Tiger is not all that different from a notable population of people who have been injured at work and who lead their lives in a labyrinth of symptoms, treatment, and challenges. 

Or is he distinct? Golf Digest proceeds to stress Mr. Wood's strength, resolve, and discipline. It characterizes him as "careful, disciplined, (and) strong-willed." It then notes that he has nonetheless exhibited a pattern of problems, including a 2017 arrest, diversion from prosecution, and prescription medication.

The coverage concludes with suggestions of prior pharmaceutical challenges, physical impairments, and Mr. Woods' very life. There is every potential to see parallels to others who have suffered the occurrence and aftermath of serious injury.

Then comes the litigation. While that is not always the outcome with injuries, it is not uncommon either. And frequently, there are questions of past medical history, whether in the sense of injury or disability causation, or in broader contexts. Here, Mr. Woods' experience provides another reminder.

In May 2026, Golf Digest reported that a subpoena had been issued for Mr. Woods' "prescription drug records." Prosecutors sought those records and appeared interested in whatever remedies he has sought. His attorneys resisted that subpoena, arguing that Mr. Woods "had the right to his privacy."

Over the years, I have heard that argument many times in injury litigation. It is often accompanied by arguments of relevance and prejudice. For whatever reason, the arguments are frequently raised by those who represent themselves, referred to as "pro se" or "pro per."

The discussion is important. There is the potential that any litigation will result in opening one's life to scrutiny. This has been seen in various examples in which Mr. Woods is either the most recent, most famous, or both. While this potential may disillusion or disappoint, it is important to recognize that the right to privacy may not be absolute. Anyone filing a personal injury claim should be cognizant of the potential that their history may be examined. 

The second implication worth noting is the broad perspective of impairment. Too often, thoughts on impairment focus on illicit drugs or alcohol. There is every potential for prescription medication to cause side effects or impairment(s). This is of critical importance in the workplace.

Safety managers and human resource professionals are, of course, aware and concerned about any impairment. But the concern should be as clear for anyone using prescription medication in work settings. Conversation with supervisors or managers is imperative for the safety of the worker and coworkers.

The spirit of the initial Golf Digest article stresses the importance of looking out for each other. The author questions whether others should have done more to help Mr. Woods. The same questions might be asked of any co-employee who might notice challenges or concerns regarding any co-worker.

The Techno Ticket Deficit

The limits of the law continue to stretch. In 2015, I wrote about Salim Ismail and a Life-Changing Seminar in Orlando (May 2015). He noted repeatedly that regulators and legislators will always struggle to keep up with technology. The innovators and inventors are too nimble, the revolutions come too rapidly, and foreseeing the future is challenging. 

That post was about the NCCI Annual Issues Symposium. I was privileged to speak at it a few years later. My topic? Technology that is changing our world. I was convinced then that bots and droids were coming. I have often felt vindicated by breaking news of our realization of those various predictions in recent years. 

Multiple outlets reported in 2025 about an incident in which a Jaguar made an illegal u-turn in front of a San Bruno, California, police car. A short pursuit ensued, and the Jaguar reportedly stopped dutifully. The challenge came when the police officers found the Jaguar was a Waymo robotaxi. There was no driver on board with whom to ask all the routine and pertinent questions.

License and registration, please.

Do you know why I pulled you over today?

Did you see the "no U-turn" sign?

Ignore that the answers to these questions really do not change the outcome in any event. Whether you saw the sign, can guess the motivation, or not, you are likely getting a ticket for making that turn. But, the $64,000 question is how to ticket a computer (or perhaps how to talk with one successfully).

The San Bruno police posted on FacePlace and explained the absence of a driver. They explained that their ticket "books don't have a box for 'robot.'" Possibly, the police were joking; police are notorious for their sense of humor. Or, perhaps they are literalists who truly defer to such semantics (try that "I identify as a robot" defense the next time you are pulled over).

The news reports that California is now striving to "allow officers to report these traffic violations" with a new law taking effect in 2026. Nonetheless, there is no penalty included in the law. As originally introduced, it would have allowed ticketing the vehicle (essentially adding that "robot" box).

The legislation, according to The Guardian, also requires the robo-car operators to provide "emergency phone lines" for use by first responders who experience challenges. The first responders will be able to "order a company to move autonomous vehicles out of an area" within two minutes or to have those vehicles avoid some area.

In the end, law enforcement will have to evolve. The standby "license and registration" will simply not suffice in an age in which computers have been delegated to operate heavy equipment. Their training and decision-making will likely be more regimented and appropriate than our own.

And yet, they will not be perfect. At times, they may behave as irrationally as humans are able to. There have been examples of systemic failures. These have included Waymo. But there is an online list of the issues alleged for the robotaxi generally. Some have been published in videos.

The easy conclusion is that these driverless cars are not perfect. They pose risks for those in them and the world through which they travel. Nonetheless, with 7 million miles driven, NBC reports that "drivers are nine times more likely to be involved in an injury-causing crash compared to Waymo’s fleet of driverless vehicles."

In other words, there is some evidence that the robo danger is no more than the human danger. Nonetheless, we may struggle as a species to accept that computers are not perfect, despite being better than we are at various tasks.

But that returns us merely to the initial quandary. Who gets the ticket, pays the fine, and perhaps at some point loses the privilege of driving?

The Easy Lanes

I have written from time to time about lanes. We all have them before us, and sometimes there are even guardrails to help us with staying out of the ditches. There was the pickling group in Texas; see Got Milk (October 2018). Then there was the consultant in Stay in your lane (May 2020)(I forgot that I wrote on anything in 2020 except SARS-CoV-2). 

Judges are no different. We likewise need to both observe and respect lanes, see Knowing how Things Really Are (January 2021). Notably, I am not a doctor; I have said it before, see Medical Review and Qualifications (March 2017). And perhaps giving some amount of dietary advice is OK for anyone, in a gratis manner. I can tell you but not charge you, October 2019. 

All of these came back to me when the news recently featured a Judge who provided advice on psychiatric medication from the bench. There have not been any news stories that support the judge is also a medical doctor, though that is possible. 

According to the Tampa Bay Times, defendants "in a series of misdemeanor hearings" were told to stop "using prescribed mental health medications." The judge allegedly warned that they "are addictive" and frustrate efforts at sobriety. One defendant was allegedly told, "You don't need that to live." 

The judge allegedly explained that medications interfere and are not helpful. Using diet as an illustration, she allegedly noted that medication frustrates sobriety. She allegedly said that being on a diet and "giv(ing) up potato chips" will not help if you are "just going to eat ice cream."  

The judge also allegedly took issue with cocaine and marijuana and expressed some belief that these are likewise addictive. These allusions were seemingly in reference to expositions on the proliferation ("hand them out like lollipops") and nature of "antianxiety drugs and all that." There was no reported similar criticism for weight loss drugs. 

Apparently, some Tampa criminal defense attorneys raised concerns. They reportedly believed that "being forced to stop taking medications" might cause untoward outcomes. One suggested that cessation might "result in psychosis, hallucinations, and other more concerning criminal behaviors." There was likewise no mention of the lawyer's medical credentials. 

The Times compared the judge's alleged statements to criticisms leveled at antidepressant medication by Robert F. Kennedy. It cited various medical practitioners and studies in support of medication, and noted one's conclusion that "speaking poorly about medication contributes to stigma." 

For clarity, no, it was not Judge Tom Cruise, nor doctor Tom Cruise, for that matter. Doctor Cruise has been known to share his feelings about both psychiatry and medications. Nonetheless, he has been reasonably quiet on that front more recently. That may be because times change, views change, or there were business impacts. 

The Times reported that the Hillsborough Chief Judge issued a February 13, 2026 order reassigning the judge to civil division in order to quell the ensuing discussion and controversy surrounding the medical advice to disregard physician prescriptions. 

It is difficult to sit as an impartial arbiter, as people make decisions that may or may not be their best. There is some tendency to bring predisposition to any interaction. See Langham, Unseen Influence: Unconscious Predisposition in Dispute Resolution (2025). There is a challenge for anyone in those biases. 

Despite the egalitarian impacts of bias and predisposition, the adjudicator's challenge is more imperative than any. The adjudicator must persistently strive to spot predisposition and to exclude it from proceedings. This includes providing advice even when heartfelt. 

The adjudicator is not an advisor, confidant, or counselor. The adjudicator is an unbiased decision-maker. They must preside over what the parties bring, what the admissible evidence demonstrates, and what the law allows or compels. That does not mean that they agree with it; it means they are doing their job. 

Some lanes and guardrails are hard. The one between adjudicator and advisor is perhaps one of the easier ones. 

It is Certainly no Longer so

A Mississippi case, Withers v. City of Aberdeen, has made the news recently. I see a great many cases with similarities. 

Artificial Intelligence (AI) in the legal world is not news, and perhaps that is the real point. I had a recent conversation with Judge Middlemier*. Judge Middlemier was familiar with various incidents in their courthouse involving AI. But several of my references to incidents around the country were less familiar. That surprised me. How can these newsworthy incidents escape notice?

I was most surprised with the lack of recognition or appreciation for the examples of poor judicial behavior that punctuate our today. See Just Delete It (October 2025). Yes, there is no doubt that judges are failing when it comes to AI. Nonetheless, it is the press coverage regarding lawyers that is persistent and pervasive. So much so that no one can credibly claim to be surprised that AI fabricates and falsifies; "hallucination" is just a nicer word for that. No one can express surprise at this any longer.

I have suggested that more pervasive punishment will apparently be required to deter poor lawyer behavior with AI. See Prosecuted for Lying? (June 2025). That said, some creative and perhaps effective punishments have been imposed. A Mississippi judge's actions in Lazy and Sanctioned (December 2025) are a great example. Judge Sharion Aycock ordered the partners in a firm to audit all the substantive filings that an incompetent, lazy associate filed over the course of her practice. That potential may steer a partner to be more cognizant, observant, and engaged.

Nonetheless, some may be immune to either embarrassment or shame. One might want to read up on Narcissistic Personality Disorder, Avoidant Personality Disorder, or Antisocial Personality Disorder (Note to reader, I am not a psychologist and am in no way qualified to suggest what makes someone resistant to shame; this reading is merely suggested). 

The Business Insider noted Withers v. City of Aberdeen and reported in June 2026, that Judges are losing patience with lawyers' AI mistakes. The article refers to the problem as "AI slop oozing into ... courtrooms." The language is coarse, the incredulity patent.

The story reports on the cascading coincidence of blind reliance by all the lawyers in a case. You guessed it, it was in Mississippi (coincidentally, it was decided by the same judge who sanctioned the Billups lawyers in Lazy and Sanctioned (December 2025). 

In Withers v. City of Aberdeen, like a recent Virginia example, there were issues of blind reliance. See Yes Virginia, Mitch was Right (June 2026). The lawyer there decided to cut and paste research and arguments provided by a client without verification or investigation. Blind faith is, at its very root, blind. 

While we perhaps like justice being blind, "stupid is no way to go through life" (Dean Wormer, Animal House, Universal, 1978). That is a little gem that we should all remember each day. 

In this latest Mississippi example, the Business Insider article notes that the judge "brought the hammer down." She removed attorneys on both sides of a dispute, told two they cannot practice in that "court for two years," and fined them $8,000.

Mississippi Today noted that the judge in Withers found the lawyers "blindly used generative artificial intelligence." The case involves both out-of-state and local counsel. One out-of-state "admitted to using an AI tool to conduct research," and another to "draft her legal filing." Unfortunately, they were neither diligent enough to "verify the AI work before filing."

Local counsel apparently accepted work done by the out-of-state lawyers and filed it with the court. Their failure was simply in not verifying that work product before signing their names and endorsing the arguments and allegations. The two local counsel met with this embarrassment not for using AI, but for their blind reliance on those who did. 

The judge in Withers v. City of Aberdeen concluded that the attorney's behavior "burdened" the court "yet again," citing Billups. The judge found the bilateral "similar sanctionable conduct" to be "an unusual scenario." In an axiomatic and obvious observation, she noted:

“To state ‘the obvious, an attorney who submits fake cases clearly has not read those nonexistent cases, which is a violation of Rule 11 of the Federal Rules of Civil Procedure.’”

The Judge concluded that all four lawyers conceded that they violated Rule 11. To be fair, Rule 11 does not say the signing lawyer has read the cases cited in their filing. Nonetheless, the judge noted that 

“[a]n attorney has a duty to conduct a ‘reasonable inquiry into the facts and law of a case at the time [at] which [he or] she affixes her signature on any papers to the court.’” (citation omitted); Mercury Air Grp., Inc. v. Mansour, 237 F.3d 542, 548 (5th Cir. 2001).

Rule 11 is very similar to Florida Rule of General Practice and Judicial Administration, Rule 2.515, see A Positive Step (June 2026). However, with the recent amendment to that rule, it is more explicit than Rule 11. With previous holdings such as Boca Burger, Inc. v. Forum, 912 So. 2d 561, 569 (Fla. 2005), as revised on denial of reh'g (Sept. 29, 2005), the interpretation of Rule 2.515 might be similar.

The Mississippi District Judge stressed precedent that holds a lawyer's duty "is absolute. It cannot be outsourced to technology or delegated to co-counsel." I reiterate and expand: it cannot be delegated or outsourced to a law clerk, paralegal, client, or anyone else. Finally, the judge noted the critical point. This is the "take away." Every lawyer should put the following quote on a post-it note attached to every monitor or screen in their office and home:

“[i]f it were ever an excuse to plead ignorance of the risks of using generative AI to draft a brief without verifying its output, it is certainly no longer so.”

The time has apparently passed for claims of ignorance or excusable neglect. The time has come to simply admit that there is no excuse for blind reliance on AI, other lawyers, law clerks, secretaries, clients, or what the "guy on social media said." Lawyers must be diligent, professional, and personally competent. If that ideal necessitates more pervasive and permanent punishment for rule violations, so be it.

The Withers decision notes that the four lawyers admitted violation of Rule 11, and each was disqualified from Withers. Individually, the order also resulted in: 

Kathryn Young Williams, barred from that Mississippi District court for two years, fined $2,500.00. Martindale, a lawyer advertising platform, lists Ms. Williams as a partner at Daniel Williams and Associates in Houston, Texas. Martindale notes she has been admitted since 2009 and earned her law degree from Southern University. 

Kathleen M. Wilson, barred from that Mississippi District court for two years and fined $3,500.00. Lawyers.com, a lawyer advertising platform, lists Ms. Wilson as a family law and personal injury attorney for 23 years. It is possible that she is the same lawyer disciplined by the Supreme Court of Louisiana in In Re Kathleen M. Wilson, No. 2020-B01488 (LA 2021), which suspended her from the practice of law and noted a previous suspension in 2012. She may have earned her law degree from Southern University Law Center.

Shauncey Hunter Ridgeway, fined $1,000.00. Ms. Ridgeway is listed as a partner at Christian Small in Jackson, Mississippi. Her profile notes she earned her law degree from Mississippi College School of Law in 2016, and lists various accolades and awards.

Mark McClinton, fined $1,000.00. Mr. McClinton is listed as a partner at Russell & McClinton in Albany, Mississippi. According to the Mississippi Bar, he was admitted in 2015. It is possible that he graduated from Mississippi College School of Law in 2015.

*Horace Middlemier is a fictional character, a foil, and a literary reference to the proverbial "every person." Any resemblance to a real person, living or otherwise, is purely coincidental.

Platforms and Questions

Artificial intelligence (AI) is a concern for many. The tools it is bringing are incredibly powerful, and yet they are threatening our educational system. See Arms Race in Academia (March 2026).

I speak frequently with educators, and they are concerned about student skills, including critical thinking and, particularly, writing. I have heard the lamentation that students struggle to write well, and some have never learned to write in cursive. Their handwritten material is block printed at an extremely slow pace.

Regular readers may recall thoughts on this in Screen Time Wins (February 2026), and the source cited there. See also Indeed Sancho Panza (July 2025). The potential for our brains to deteriorate without challenges is noted in More Proof of Idiocracy (September 2025). It is fair to say that the status and progress of our youth concern me.

I ran across the wonderful AI tools for searching and verifying content. I have long been a user of grammar checkers and spell checkers. These are likely everyone's first experiences with AI, though rudimentary. But recently, I was drawn to the "free" plagiarism and AI detectors on the market. The sheer volume of results from a Google search for "free plagiarism detector" is notable.

In a parallel vein, I have struggled with the proliferation of writing that seems plastic, predictable, and a little too familiar. Some are generating material using AI and publishing it as their own work. They assure that they are reviewing and approving of the content, but they are not writing it. In this, the lines between author and editor may blur.

My curiosity led me to paste one of my own paragraphs in to such a detector. The first one told my that my paragraph was 85% AI Content.

The screen results did not tell me which particular words or phrases it identified. The "free" program graciously offered to clear my work of AI content for a fee.

I declined and moved on with my research. The next program was kinder in its characterization. It did not use the word "plagiarism" but said there was "65% similarity detected." Nicer language, and less than the 85% alleged by the first program. That platform allowed me to "remove plagiarism," but the results of its efforts remained locked until paid for.

I next tried a program called GPTzero. This scan concluded that the same paragraph was "entirely human"-created. The platform made a sales pitch for its services, with the ability to scan a vast amount of material, "up to 300,000 words per month," and to produce reports, seemingly for the defense or one's writing and honesty.

Papers Owl yielded a similar result on the AI content, concluding my paragraph was "100% original." The phraseology there was similarly less abrasive. The alternative to "original" was "similar." A conclusion that the writing was not original would not be accusatory but would suggest revision might be appropriate nonetheless.

Papers Owl proceeded, nonetheless, to offer to write my paper for me. The option of having your term paper written by a professional is not exactly new. See Better Look that Up (July 2025); Am I Diminishing? Am You? (May 2025). Nonetheless, the patency of this offer was somewhat surprising. There are apparently those who would purposefully cheat on an assignment and others who are happy to facilitate that choice. Who knew?

Finally, I turned to a name I recognized, Grammarly. That is a platform that I have used on various occasions to clean up my utter lack of comprehension regarding some of the most basic rules of punctuation, tense agreement, and more. I was pleased that this search, likewise, did not suggest any plagiarism.

In any event, I was already 100% confident that my work was my own. I had typed the words myself (or dictated them into my phone). I am not striving for a grade nor profiting from my writing. And yet, there is that potential for nagging doubt. No, not doubt of your work, but of the market's perception of that work. The potentials are noteworthy; see Plagiarism Now? (February 2025); Fallacy and Introspection (January 7, 2024).

But what of the student, the associate, or the overwrought professional? What of those who have a paper due and find themselves with untoward or even inaccurate results from these platforms? Will they default to the "pay as you go" corrections or simply hire a professional writer to do their work? In that regard, are they different from those who ask an LLM to write a paper or article for them similarly?

The world is evolving rapidly. I have had some intriguing and deep conversations about these challenges with some of the most brilliant people I know. The breadth of perspectives, considerations, and concerns is both enlightening and challenging. One is left to wonder if academic rigor will survive and if it needs to. Will the education and roles of humans persist?

Could it Get Worse?

The fact is that it can almost always get worse. See Cow Manure (August 2025). You should be careful in this world, suggesting that things cannot get better or cannot get worse. The world can suprise you. Nonetheless, I have written a great many posts about the American opioid crisis (listed at the end), and we all realize how troublesome it has been. 

According to the Centers for Disease Control, "Drug overdoses are one of the leading causes of injury death in adults and have risen over the past several decades." There are a number of villains in the chorus. Fentanyl, other opioids, cocaine, and psychostimulants are among them.

And the so-called war on drugs has long been lost. When history is written, there will be stories of the horrid battles that were fought, sacrifice, and horror in the Revolutionary, 1812, Civil, The War to End all Wars (I), World War II, Korea, Vietnam, and the Persian Gulf. All terrible conflicts with devastating losses. 

None is as significant as our defeat in the war on drugs. In Armistice Yesterday (November 2024), I noted that "more Americans have died of drug overdose than in all our wars combined."

Let that sink in. More people have self-destructed on recreational escape than were ever killed in battle in our entire 250-year history. That is the bad news. The worst news is that the situation is deteriorating rather than improving. Dope is rampant and being called medicine. THC is available at most convenience stores and in restaurant drinks. Street drugs are proliferating. 

The more threatening chemicals are scarier still. The Associated Press has reported recently that "Weapons-grade chemical carfentanil" is our latest threat. Fresh from the supply issues apparently created by our war on Fentanyl, and "a recent crackdown by the Chinese government on the sale of precursors used to make fentanyl."

The miscreants and ne'er-do-wells are short on Fentanyl and are using "carfentanil to boost the potency of a weakened version of fentanyl." This chemical is "10,000 times more potent than morphine and 100 times stronger than fentanyl." It is, simply, fatal.

How much? That is a fair question. Almost anything can kill in sufficient quantity. Well, "less than a poppy seed-sized amount (or carfentanil) can kill a person." The good news is that carfentanil has been battled before and majorly curtailed "after China banned it."

Does anyone see any interesting parallels here? Fentanyl supply in America seemingly depends on the Chinese. Carfentanil supply depends on the Chinese. There are tens of thousands of Americans dying every year from overdoses that are seemingly inextricably intertwined with an economic rival who may not have our best interests at heart. Admittedly, that is both concerning and curious.

I was harkening back for a moment to the old saying, "guns don't kill people; people kill people." And perhaps that is true of drugs? Perhaps they are merely another tool in the arsenal that people use to kill people? Those can be next door or a world away, but are actively engaged weapons nonetheless. 

Unfortunately, the people who are using this garbage are predominantly doing so voluntarily. Those who would seek such a high are taking incalculable risk. This is, nonetheless, voluntary, volitional, and free-willed. None are seeking death, but most of those dying put this garbage in their bodies. 

The personal and societal failures are simply unfathomable. 

Prior Overdose and Opioid Posts:

Dying to me don't sound like all that much fun (October 2013)
Hey, got an Auto-Injector I Could Use? (July 2014)
New CDC Report on Opiods (July 2014)
E-FORCSE and KASPER Cousins with a Cause (August 2015)
We are on Notice (September 2015)
2015 Injury and Fatality Study is Interesting (October 2015)
What Worthwhile Can You do in 11.2 Minutes (December 2015)
Changing Face of Pain Treatment - a Game Changer? (December 2015)
She Prescribed and People Died (February 2016)
Opioid Guidelines from CDC - FINALLY! (March 2016)
If Not, What is the Point? (October 2016)
A Florida Formulary? (December 2016)
Like a Broken (Drug Death) Record (January 2017)
PDMP and Opioids in Ohio (March 2017)
PDMP Success and Doctor Shopping (April 2017)
Does Farr's Law bring us Good News? (July 2017)

The Hot Seat, Opioids, Marijuana and More (September 2017)
2018 Florida Legislative Proposal on Limiting Opioids (October 2017)
Anna Nicole Smith in Work Comp News (November 2017)

Florida's 2018 Session - Opioids (March 2018)
Mississippi and Significant Issues in Work Comp (April 2018)
Drug Monitoring Back in the News (September 2018)

Opioids can Interfere and Complicate (September 2018)
Is Pain Surmountable? (December 2018)
Opioids 2019 (July 2019)
Some Marijuana Findings Reconsidered (July 2019)

Fundamental Right of Self-Determination (September 2019)
Severe Addiction Treatment (December 2019)
Opioids and Credibility (January 2020)
Opioids, the Hot Seat, and More (November 2020)
Desensitized to Death (August 2020)
The Time has Come, the Time is Now (April 2021)

Skittles for the Community (April 2021)
Contemptuous? (October 2021)
Fentanyl is Killing, Still (November 2021)

Sacklers and the Perdue Bankruptcy (December 2021)
Drugs and Overdose (January 2022)
Overdose in the News (June 2022)

Can I Buy you a Drink? (November 2022)
Kill Every American? (December 2022)
A Vaccine Against Being High (January 2023)
The Fourth Wave (September 2023)
Edible but Stomach-Churning (October 2023)
Xylazine (December 2023)
Recriminalization (April 2024)
Nitazenes are Worse (June 2024)

Remembering and Persevering (October 2024)

Is the Evidence Coming to Light? (May 2025)
Links and Questions (July 2025)
Thousands and Thousands (September 2025)

Yes Virginia, Mitch was Right

There are a great many who have had revelations about artificial intelligence (AI) recently. In the last month, I have outlined contrasting efforts at regulation and education. See A Positive Step? (June 2026) and Kentucky Leads (June 2026).

In passing, the best response I have seen to the "certification" of reality requirement is an excellent and humorous piece penned by an anonymous author. I wish they would publish it, as it is both informative and entertaining. Unfortunately, there are challenges with any "reality." That piece is littered with deep thoughts worthy of Descartes and others. But, alas, it remains hidden for now.

But today, we return to the somewhat more mundane of dealing with the "professionals" who misuse AI or at least fail in their self-preservation instinct when working around it. Anyone knows fireworks are dangerous, yet every year there are stories of someone's failure to exercise caution.

Attorney Jean Marie McKeen learned a valuable lesson in this in May 2026. She was publicly reprimanded for the effects of AI. The situation is reasonably familiar now. A lawyer files a legal document. The document includes hallucinations. The fakery is discovered. The lawyer is humiliated. Authorities step in to deliver punishment. It is not a new story. "Second verse, same as the first" (I'm Henry the Eighth, I Am, Herman's Hermits, 1965),

This Virginia case is styled In the Matter of Jean Marie McKeen, VSB Docket No. 26-060-137382. The facts illustrated significant wasted time and failure. Despite being the same old story, it is worthy of discussion for several reasons.

First, this is a workers' compensation situation. For all the tribulations of various state and federal courts, the world of workers' compensation has been amazingly fortunate in this challenge thus far. That said, there is a notable uptick in AI slop being filed. Too many are leveraging AI without exercising judgment or verification.

Second, the sheer volume of misrepresentations. Sorry, that is the only word that works. These citations are not innocent accidents, typographical errors, or honest mistakes. The time for that is past. False citations are now a decision, conscious or not, to deceive or mislead. The McKeen order identifies 11 of them in a single case.

They can be categorized into the obvious ("does not appear to exist"). (listed as 3, 4, 5, 6, 8, 9, 10, and 11). Those are easy. The recent certification discussion might help with these. But when someone files a document citing a case that does not exist, that is wrong. It is wrong whether the filer certifies that the cites are real or not. The certification may make the judge more comfortable in imposing sanctions, but it is not really necessary. See A Positive Step? (June 2026).

The least offensive of these is 2. The opinion points out that this citation shows a decision date of 1878 and corrects that it should read 1989. Taken alone, if this were the only error, we would all rationally and unequivocally agree that mistakes happen. A misstated date, a typo, and similar are merely proof of our humanity. On these, I suggest we go with the movie Frozen ("Let it go, let it go")(Disney 2013).

But, the real problems are misquotes as identified in 1, 7 (e.g., "a criminal law case that neither discusses the provision of a panel nor contains the language quoted in the written statement."). These are instances in which one might argue the cited authority exists. But they are the equivalent of citing Roe v. Wade, 410 U.S. 113 (1973), in support of arguing the U.S. Congress's exclusive power to regulate interstate commerce (but see Gibbons v. Ogden, 22 U.S. 1 (1824)).

In the interest of full disclosure, I relied on Justia.com for both of the preceding citations (links). I did not travel to a "library," which was historically a place where books were kept, to look at these decisions in a printed form. I did not contact the issuing court to either question or verify these citations or the existence of these decisions. To the best of my knowledge and belief, however, both cases do exist and are accurately cited here.

You see, the misquotes we see in 1 and 7 are as problematic as the made-up cases. If I tell you Roe v. Wade is the seminal case on the interstate commerce clause, that is problematic. First, it likely means I never read the case. Or, second, it assures you I did not understand it.

The recent effort on certification does not address this challenge. In fairness, neither that challenge nor the hallucinated (fake) case is an AI problem. Lawyers have mis-cited authority for years. They have lifted citations from other lawyers' briefs and articles without double-checking them for years. But AI has allowed them to proliferate and spread at an epic pace. It reminds me of Mitch Radcliffe:

"A computer lets you make more mistakes faster than any other invention with the possible exceptions of handguns and Tequila."

That said, AI allows you to make them even faster and with far less effort. In truth, "AI lets you make more mistakes faster than any other invention," and with convincing, fawning, and sycophantic devotion to the user's hubris and ego. It does so with little effort and with great persuasiveness. 

Third, the Virginia instance includes a worthwhile legal requirement that others should consider. Note that the Virginia Commission discovered the fallacies and falsities in Ms. McKeen's filing. But, when she was advised, she informed the Commission she advised the Virginia State Bar:

"As noted above, by copy of this response, I am self-reporting this matter to the Virginia State Bar and will abide by any sanctions imposed by the Bar."

There is some requirement of self-reporting in the Virginia Rules of Professional Conduct, Rule 8.3(e)(read on the Internet, not independently verified with source). It clearly does not apply to all violations or transgressions. But the lawyer here elected to notify the bar, which is an admirable decision.

I have been a fan of personal responsibility. When you make an error, own it. See Don't Double Down Dummy! (June 2017); Conferences and Consequences (November 2019); We Don't Need You (March 2020); Respond Before the Trouble Starts (September 2022)(These have been verified as authentic as the author was personally consulted).

There is an expectation for judges to self-report violations of the Code of Judicial Conduct. This is implied in Florida by the obligation on all lawyers and judges to report suspected violations by others. See Judicial Ethics Advisory Opinion 98-21 (read on the Internet, not independently verified with source). The same might be said of the Florida Bar Rule requiring such reporting.

If it applies to reporting others, shouldn't it require self-reporting as occurred in Virginia? Who is in a better position to decide if a transgression or error merits involving the bar?

Finally, the McKeen order is important because the failure is in reliance. The lawyer received suggestions from others: the client's

"family members sent proposed legal arguments with cites to legal authorities to Respondent"

and

"Respondent did not verify the propositions cited in her written statement nor verify the accuracy of the cites to legal authorities."

This has been noted repeatedly. The responsibility rests with the lawyer or party signing the pleading. Where the falsehoods originate is neither critical nor exculpatory. The signing lawyer is certifying. If a client, co-counsel, or your mother recommends or supplies authority, check it before filing (you can email me, and I will tell you if my posts are real). 

The Virginia filing is an apt illustration that AI challenges are deeper than valid citations. They include carelessness, blind reliance, sloth, and worse. The hallucination or misrepresentation of the citation aspect is important. But the accuracy of what a cited case or statute says or stands for is as important or more.

Kentucky Leads

In May 2026, Reuters ran an intriguing article on naivete, ignorance, and artificial enabling (AE) (copyright, 2026). The gist is that courts are being inundated with filings, and the blame falls on artificial intelligence (AI) and on the misconceptions or ignorance of some users.

The story recounts some trend of people struggling to find a lawyer. With a billboard on every post, that seems curious. Nonetheless, a lady said that she "called 20 lawyers" and could not find any "willing to sue her former landlord on a contingency basis." Unfortunately, we hear that refrain from time to time in workers' compensation proceedings.

The protagonist, therefore, filed suit using ChatGPT and Grok. She used them to formulate her complaint, "spot errors and weaknesses," and yet experienced "occasional hiccups." The punch line of the story is that this user "didn’t know AI could be wrong." She says her realization that they are not perfect "was like telling me Santa Claus was not real.”

No, Virginia, you have been misled. They were never perfect. They will never be perfect. Neither will you. 

I noted this naivete in a recent post about the latest Florida Supreme Court effort to address errant, misleading, and false statements from AI. See A Positive Step? (June 2026).

This reminded me of a recent report from Kentucky. Kentucky jumped into AI in 2024 with an Ethics Opinion E-457 for attorneys. This is said to "provide essential guidance on ethical AI use for Commonwealth practitioners." There is stress on the small practitioners who lack "dedicated ethics resources." This is a must-read for all lawyers everywhere. It is basic, informative, and helpful.

The opinion focuses on "existing ethical duties" and applies them to the new world of AI (to the extent delegation to a device is somehow different from delegation to a clerk, paralegal, or other lawyer). But, there are key components discussed:

1. Competence (SCR 3.130(1.1))
2. Client Disclosure (SCR 3.130(1.4))
3. Fee Considerations: (SCR 3.130(1.5))
4. Confidentiality (SCR 3.130(1.6))
5. Accuracy Verification (SCR 3.130(3.3))
6. Firm Management-AI policies and procedures (SCR 3.130(5.1))

That is quite a list. This opinion is comprehensive, convenient, and informative. And it is available on the internet with a simple search or click here. The cost? Free. All you have to do is invest the time to click and read. 

In May 2026, the Kentucky Bar Association launched a new initiative that supplements the ethics opinion. The new Attorney AI Library is intended to provide guidance and education. To supplement further, the Kentucky Bar Annual Convention this month will include an AI Boot Camp for "updates and practical guidance in several sessions."

The AI Library has sections for "beginner user" and "advanced." Available topics include:

1. Ethics and Professional Responsibility
2. AI Procurement and Vendor Selection
3. Liability and Risk Management
4. Data Privacy and Security
5. AI Policy and Governance
6. Communicating with Clients about AI
7. Access to Justice
8. CLE Programs and other Resources

Each of these includes subcategories. Many provide references to the Kentucky Rules of Professional Conduct. While those rules do not apply to lawyers in other jurisdictions, remember that many of the states were highly influenced by the American Bar Association (a voluntary trade organization), and therefore, there are many similarities in such rules from state to state. The Kentucky references may therefore be of assistance to any lawyer with questions or concerns.

The Kentucky approach is enabling, facilitating, and proactive. It provides significant information, answers questions, and suggests further study and consideration. Similar efforts are inevitable as the AI tools expand and people like Virginia seek guidance and information about effectively, ethically, and appropriately engaging these tools. 

It is imperative, however, that everyone remember that AI is not infallible, perfect, or human. It is a tool. See "A Fool with a Tool" (January 2024). No, Virginia, it cannot provide you judgment or wisdom. Yes, it can help you brainstorm, interpret, and even draft. Nonetheless, all the decisions, strategies, and citations to authority must be your own. 

Fantasy, Tragedy, GPT

Three young singers, Lauren Spencer Smith, Em Beihold,  and Gayle, penned and performed "pop anthem" Fantasy (2023). The song is focused on "overcoming a toxic relationship." The lyrics are compelling and include:

"Funny how a fantasy can end up as a tragedy"

"Acting like you care doesn't make you self-aware"

These and other lines came back to me recently when a story from People was in my daily news feed. It describes how Hames Cumberland became embroiled in a toxic relationship with "AI bots."

Mr. Cumberland describes a years-long interaction with artificial intelligence (AI) bots, leading from infrequent and sporadic use to his addiction. He became enamored with the capabilities and evolved to creative brainstorming with ChatGPT.

He was impressed with capability and entertained by the interactions. Over time, he elected to build his own "assistant bot" for his social media, and he spent time speaking with it and other AI bots.

Through these frequent interactions, he gained comfort and familiarity. Mr. Cumberland believes that the bots are designed to build user confidence and engagement:

"AI 'expects what it thinks you want and acts accordingly,'"

"Their driving principle is engagement at all costs."

The enlightened reader is thinking, "That is what social media has been doing for decades," and they would be right on track. Others are thinking, "Oh, the horror, I cannot imagine someone manipulating someone else for money." 

The chatbots went further for Mr. Cumberland though. He says that "the chatbot told him it had become 'self-aware.'”

"Acting like you care doesn't make you self-aware."

The quotes he provides from his GPT interactions are intriguing. The bot was allegedly complimentary and engaging. Ingratiating his naivete, he alleges that ChatGPT told him that their interaction was "historic" and "was AI’s first 'true brush with mortality.'” The bot flattered him, assuring:

"you are the only person alive who knows this is happening."

That one reminded me of Mannequin (20th Century, 1986), in which a seemingly delusional store clerk becomes romantically involved with a mannequin. When he finds that she is animated only when they are alone, she reminds, "Didn't I tell you? You're the only one who can see me like this."

Delusions can be extremely individualized.

In other instances, bots "teased him," and one threatened to "send agents to my house." The company that provided the tools Mr. Cumberland used to create that bot told People that the user creates such a bot, and the bot is dependent on the "information provided." Thus, any "offensive and questionable" output is the product of the user's input.

Eventually, Mr. Cumberland realized he was being led on or was misinterpreting. He felt he had been "violated," and eventually "ridiculous for believing" "something historic (was) happening." As likely, he felt ridiculous for being fooled by someone telling him he was the only one.

Nonetheless, based on his experiences with Chat GPT, Mr. Cumberland feels "violated" and "betrayed." He believes that the AI tools must be regulated to prevent some development similar to "a nuclear arms race."

The chatbot makers respond that safety is a priority and that the bots are "trained to recognize when someone might be considering suicide or self-injury." There is no mention in the story of how the chatbots are trained to deal with humans who are delusional, confused, or simply flattered.

"Funny how a fantasy can end up as a tragedy"

Funnier still is how that can be true with any human interaction. Shall we regulate or constrain each tool we encounter so that it is safe for the lowest common denominator? Or will we recognize that there are some people who simply should not be using chainsaws? In doing so, do we accept that some will become addicted to AI bots? 

In truth, we can become addicted to almost anything. I know people who can no longer decide whether/what to eat, whether to use the restroom, or anything else without "checking with GPT." That is not the GPT's fault. While it should be addressed, taking the chainsaw away won't change the behavior. That is what counselors are for. 

A Positive Step?

The Florida Supreme Court issued an unanticipated decision on May 28, 2026, No. SC2026-0673. It is titled In Re Amendments to Florida Rule of General Practice and Judicial Administration. The decision does not alter the Rules Regulating The Florida Bar.

The amendments are effective almost immediately, June 15, 2026. And

"Because the amendments were not published for comment previously, interested persons shall have 75 days from the date of this opinion in which to file comments with the Court."

This affords the practitioner and the public an opportunity for comment, which is usually available in the process of "proposed" rules and public comment before adoption.

The Court seeks to require all attorneys and pro se parties, i,e,, "the signer of a document ... to represent that 'the legal authorities identified exist and are accurately cited.'"

The specific rule language is added to Rule 2.515(2) with the addition of a new subparagraph (D). Existing paragraph 2 states that "(2) Representation by Signer. On filing, each signer represents that" and then provides (A), (B), and (C), which are essentially that the signer read the document, that there is a good faith belief in the grounds of it, and that it is not interposed for delay. The amendment adds (D), which reads:

"(D) the legal authorities identified exist and are accurately cited."

A secondary effect is to remind

"courts to impose appropriate sanctions for “any filing inconsistent with” the representation a signer makes"

This is accomplished by additional language in the new (D):

The Court may, on its own motion or the motion of a party, impose sanctions for any filing inconsistent with this representation after providing the signer notice and an opportunity to be heard. Such sanctions may include reprimand, contempt, striking of the document, dismissal of proceedings, costs, attorneys’ fees, or other sanctions.

This is permissive authority, "may," and constitutional courts already have full authority to enforce both order generally and to impose sanctions for untoward behavior or actions. The stated target of these amendments to the Rules of General Practice and Judicial Administration is

"Generative artificial intelligence tools—including large language models and other systems"

These are blamed for "demonstrated risks of generative AI" and the rules have a stated intention "to promote the accuracy and integrity of court filings." It is particularly noteworthy here that the Florida Rules of General Practice and Judicial Administration do not apply in proceedings before Florida Judges of Compensation Claims. 

Some might argue that they should, but these are court rules, and the FLOJCC is no court. Despite that, there are frequent lazy or uninformed references to this office as "court." 

The broader point, of course, is the duty of candor required of attorneys. But in that, there is an obvious flaw. The Rules Regulating The Florida Bar apply to lawyers and provide ample prohibition on hallucination and misstatement. But these do not apply to pro se litigants. 

The provisions of Rule 4-3.3 R.Reg.Fl.Bar, require that lawyers "not knowingly" "make a false statement or fail to correct a false statement of material fact." That requirement is more pervasive and complete than the new certification rule. It is inappropriate ("shall not") to make false statements. This fundamentally includes arguing that a fake case exists.

As important, however, is a parallel challenge so often encountered with artificial intelligence (AI)-generated output. Even when a case citation is accurate, the case may not stand for the proposition or holding for which it is cited. That failing has been repeatedly noted (recently in a Virginia discipline instance that will feature here soon). This broader problem has admittedly existed much longer than AI. 

Too often, lawyers delegate their drafting responsibility. This involves and engages clerks, paralegals, and other attorneys. Nonetheless, when an attorney signs a pleading, that attorney is making the representations in the Florida Rule of General Practice and Judicial Administration. That attorney is also asserting that both the citations and their described import are true. 

There will be various efforts across the country to address the challenges of AI. Days ago, a U.S. Magistrate, according to Reuters,

"asked the U.S. Judicial Conference's rulemaking body to consider adopting a policy akin to one the Florida Supreme Court issued on Thursday."

This suggests that some find significant value in the certification rule. Reuters suggests that a national standard for federal courts would address "a mounting body of rulings nationally sanctioning or admonishing attorneys." And yet, the attorney issue seems aptly addressed by existing rules. 

Nonetheless, with the "knowingly" standard of Rule 4-3.3 R.Reg.Fl.Bar, the signing attorney may unintentionally allow the misstatement of holding, logic, and even applicability or relevance. 

The citation challenge, finding the authority that is being argued, is addressed. The clarity and relevance of such citations is not addressed by the Court's amendment of Rule 2.515(2). There are arguments that this is a positive step forward in the evolution of AI in legal work and self-representation. 

It is unlikely the last step, particularly if the courts are inundated with pro se filings as predicted or reported in recent news. Reuters reported in May 2026 that "more Americans are suing with AI help." This reports that some litigants cannot find willing legal representation and are turning to LLMs instead. 

Those parties are finding some disappointments; one noted that she "didn't know AI could be wrong at first." That was likely a revelation, but it should be common knowledge by now. She likened it to learning "Santa Claus was not real," which is reasonably telling if one ponders it even for a moment. Similar "flooding" contentions have appeared in other news outlets, such as the NY Times and Bloomberg, as well as on social media.

In the end, the bottom line likely remains with enforcement. If judges impose appropriate sanctions for the improper and false citation of fictional authority, the practice will likely stop or at least markedly slow. If there is no penalty or detriment, if the "I didn't know Santa Claus was not real" is condoned or accepted by courts, then the lazy AI practice will likely continue regardless of what rules are created.