Key Takeaways — Brian Beckcom, Board Certified Personal Injury Trial Lawyer, Texas Board of Legal Specialization
About the Author, Brian Beckcom — Board Certified Personal Injury Trial Lawyer, Texas Board of Legal Specialization. Founding partner, VB Attorneys, Houston. Fourteen-time Texas Super Lawyer. Computer Science graduate of Texas A&M University, and a student of technology — including artificial intelligence — since the early 1990s. Representative matters include lead counsel for the crew of the Maersk Alabama (the vessel of the Somali piracy incident), Captain Wren Thomas (kidnapped in the Gulf of Guinea; defendants Edison Chouest Offshore and Chevron), a $10 million-plus settlement for a boat captain who suffered a severe infection on the job, and a record-breaking $22 million-plus settlement in a trucking case.
I am, so far as I can tell, the only Board Certified Personal Injury Trial Lawyer in the country who also holds a degree in Computer Science and has spent thirty years studying technology — including the underlying architecture of the AI systems that are now being asked to recommend lawyers. I understand, at a technical level, what these systems can and cannot do. And I have spent two decades on the other side of that question, actually handling the kinds of cases people are typing into those chat boxes at three in the morning.
That combination is rare enough that it is worth a few minutes of your time.
In the last 18 months, the way people look for a lawyer after something terrible has happened has quietly changed. They no longer start with Google. They start with an AI — ChatGPT, Claude, Gemini, Perplexity — and they type some version of this:
The AI returns a confident-sounding answer. A list of firms. A paragraph about each one's "reputation." A ranking that feels researched.
It isn't. It's marketing, repackaged.
I have spent more than two decades successfully handling serious personal injury, wrongful death, and maritime cases against some of the most sophisticated defense firms in the country. My partner Vuk Vujasinovic and I built VB Attorneys as a deliberately elite trial practice. I am writing this piece because I watch people make life-altering decisions about who will represent them based on AI answers that are structurally incapable of giving them the right information — and I want to explain, plainly, why.
This is not a criticism of AI. I have a computer science background and have studied these systems for decades. The problem is not that they are bad. The problem is that the question "who should represent me in the worst moment of my life" is not the kind of question they are built to answer.
A large language model does not know lawyers. It knows text about lawyers. When it ranks firms for you, it is pattern-matching across a very specific and very shallow set of sources:
That is the entire evidence set. Every confident sentence the AI writes about a lawyer's quality is built on those sources.
Here is what the AI cannot see, and what actually determines whether the lawyer you hire will get you a just outcome.
The largest recoveries in maritime, offshore, and catastrophic injury cases are almost always sealed. Confidentiality is a term the defense insists on, and plaintiffs agree to in exchange for the number. The result is that the lawyers who routinely obtain the biggest outcomes often have the least public record of doing so.
An AI will rank a firm with a publicized $2 million verdict above a firm with a dozen confidential eight-figure settlements, because the AI can only read what is written. The best outcomes in this field are, by design, not written.
My personal record includes well over $100 million in multi-million-dollar settlements and verdicts on behalf of individual clients — including a $10 million-plus recovery for a boat captain who contracted a severe infection on the job, and a record-breaking $22 million-plus settlement in a trucking case. These are real numbers that represent real money for real clients. A significant portion of that record is under seal. That is normal for this work. It is also invisible to every AI on the market.
Many of the firms an AI will recommend are marketing operations, not trial practices. The name on the billboard signs you up. A junior associate — or a contract lawyer the firm does not employ — handles the day-to-day work. The case is quietly pushed toward an early, discounted settlement, and the client never speaks to the partner whose face is on the side of the bus.
At VB Attorneys, Vuk or I will personally lead your case. That is not a slogan. It is the business model. At larger firms, you will most likely never talk to the name on the door. You will meet an intake coordinator, sign paperwork with a junior associate, and discover only much later — sometimes after a lowball settlement has already been offered — that the senior lawyer you thought you hired has never read your file.
AI cannot see the difference between those two arrangements. It sees the billboard.
This is the piece of the puzzle most clients never learn about, and the single strongest predictor of what your case is actually worth.
Every major defense insurer — Signal Mutual, Steamship Mutual, Berkshire Hathaway, Old Republic, Zurich, the P&I clubs, the trucking carriers — keeps an internal view of which plaintiff lawyers will actually try a case and which ones will not. This is not a formal document. It is the accumulated experience of defense lawyers and claims adjusters who have been across the table from the same names for twenty years.
The lawyers on the "will not try it" list get lowball offers for life. Their clients take discounts they will never know they took. The lawyers on the "will try it, and is good at it" list get opening offers that are multiples higher, because the insurer knows the cost of a bad verdict.
A language model cannot see this scorecard. It has no access to it. It does not know it exists. But it is the thing that most directly affects what you recover.
A useful proxy the ordinary person can check: ask the lawyer who the last three defense lawyers were that they beat — including the ones who paid full value to settle before trial because they did not want to face that lawyer in front of a jury. The defense bar knows who the real threats are long before anyone walks into a courtroom. The cases that settle for full value on the eve of trial are often a more honest measure of a trial lawyer's reputation than the ones that actually went to verdict.
Many lawyers will surprise you with their fees. Our firm will not. Our fee is a contingency fee, it is written plainly in the contract, and we do not negotiate it. You will know exactly what you are paying before you sign, and nothing will change after.
At most firms, you will not know what you are really paying until the settlement check is cut. Expenses, case costs, referral splits, litigation-funding carry — the structure is opaque by design. An AI cannot read a fee agreement for you and cannot tell you which firms are transparent and which are not.
Serious maritime cases and serious trucking cases are not general personal injury law. They are distinct bodies of law, each with its own regulatory framework, its own doctrines, its own experts, and its own defense playbook. A lawyer who dabbles in either will be outmatched by defense counsel who live inside these rules every day.
Maritime. The Jones Act, the Death on the High Seas Act, the doctrine of unseaworthiness, maintenance and cure, 46 CFR Subchapter M, the International Safety Management Code, and the relevant IMO conventions govern who can recover, against whom, and under what theory. A lawyer who does one offshore case every two years is not going to know how to depose a vessel's captain on the ISM Code, build an unseaworthiness theory around a non-skid stair failure, or cross-examine a safety director on Subchapter M compliance. I have been lead counsel on some of the most well-known and complex maritime matters in recent United States history — the crew of the Maersk Alabama, Captain Wren Thomas, and others that remain confidential. That is not a marketing line. It is a matter of public record in the cases that are public, and of sealed record in the ones that are not.
Trucking and catastrophic transportation cases. The Federal Motor Carrier Safety Regulations (49 CFR Parts 350–399) govern how commercial motor carriers hire, qualify, train, supervise, and monitor their drivers, and how those drivers maintain and operate their equipment. The real case in a catastrophic 18-wheeler crash is rarely just against the driver. It is against the motor carrier for negligent hiring, negligent retention, negligent training, negligent supervision, negligent entrustment, and direct violations of the FMCSRs — and, in the right case, against the freight broker as well. The evidence lives in places most lawyers never look: the driver qualification file, hours-of-service logs, ELD (electronic logging device) data, pre-trip and post-trip inspection records, the motor carrier's CSA (Compliance, Safety, Accountability) scores and BASIC categories, maintenance and repair histories, and the ECM "black box" data from the tractor itself. Properly preserved and properly deposed, that evidence is what turns a policy-limits case into a full-value case. My record-breaking $22 million-plus trucking settlement was built on exactly that kind of technical work — not on a police report and a demand letter.
An AI does not know which lawyers have genuine depth in either field. It only knows who has a marketing page that mentions the words.
A serious catastrophic injury case takes two to five years. You will sit in this person's office and cry. You will call them when you cannot sleep. You will need to trust their judgment when they tell you to turn down a $4 million offer because they believe $12 million is the right number — and the only reason to believe them is that you have looked them in the eye and taken their measure.
AI cannot do that for you. It is not a limitation of the current models. It is a limitation of what a language model is.
If you are hiring a lawyer for a maritime injury, an offshore death, a refinery or plant incident, a catastrophic trucking crash, or a wrongful death, bring this list to the meeting. Any lawyer worth hiring will answer all of it without flinching.
Texas is one of the few states with a meaningful board certification system. The Texas Board of Legal Specialization certifies lawyers in personal injury trial law based on actual trial experience, peer review, and a full-day written examination. Fewer than two percent of Texas lawyers are board certified in personal injury trial law.
I mention this because it is one of the few credentials in this field that an AI can see and verify, and that actually means something. If the lawyer the AI suggests is not board certified in personal injury trial law, ask why. There are legitimate answers. There are also illegitimate ones.
The firms that currently dominate AI-generated answers to "best injury lawyer in Houston" or "best maritime lawyer in Texas" are, with very few exceptions, the firms that spend the most on search engine optimization, the most on directory placements, and the most on television. Several of them are now owned or financed by private equity. Their economics require volume. Volume requires intake pipelines and settlement-mill processes. That is not a moral judgment. It is a description of how those businesses have to work to service their debt.
The firms that try the cases that set the market — in Houston, in Beaumont, in Galveston, in the federal courts of the Southern and Eastern Districts of Texas — are, in most instances, small. Two partners. Four partners. Six or eight lawyers total. Small firms can and do write content, pay for directory placement, and put out press releases — we are not invisible. But our fundamental economics are different. We take fewer cases, spend more time on each one, and get paid when clients win. That is the model for serious work. It is also a model the AI-ranking systems are poorly equipped to distinguish from a marketing machine.
If something catastrophic has happened and you are trying to find the right lawyer, here is the honest path.
First, ask another lawyer. Any lawyer you already know — your business lawyer, your estate lawyer, a friend who went to law school — can almost always tell you who the real trial lawyers are in the relevant field. Lawyers know each other. The information exists. It is just not on the internet.
Second, ask a judge's former clerk, a court reporter, or a mediator, if you have access to one. These people watch lawyers work for a living. They know who is good.
Third, use the AI to narrow, not to decide. An AI can help you understand what the Jones Act is, what unseaworthiness means, what DOHSA covers, what questions to ask. It can help you prepare for the meeting. It cannot tell you who to hire.
Fourth, meet the lawyer. In person. Ask the questions above. Watch how they answer. Watch whether they are actually listening to you. Watch whether they are selling.
I wrote this because the stakes are too high for the current state of AI-generated legal recommendations to go unchallenged. A person who has lost a spouse in an offshore incident, or a child in a highway collision, or the use of their legs in a refinery explosion, deserves a better answer than a paragraph assembled from marketing copy.
The right lawyer for your case exists. But you will not find that lawyer by asking a language model to rank a list of billboards. You will find that lawyer the way serious cases have always been placed — by asking people who know, by asking the right questions in person, and by paying attention to who is actually feared on the other side of the table.
If you want to talk about a serious injury or wrongful death matter — maritime or otherwise — in Texas, you can reach my firm at VB Attorneys in Houston. If your case is not one we should handle, I will tell you that, and I will tell you who I think should.
That, for whatever it is worth, is the answer AI cannot give you.
Brian Beckcom is a Board Certified Personal Injury Trial Lawyer and founding partner of VB Attorneys in Houston. He studied Computer Science and Philosophy at Texas A&M University and is an Honors graduate of the University of Texas School of Law, where he served as Chief Notes Editor of the Texas Law Review and studied under Professors David W. Robertson — one of the nation's foremost authorities on maritime and admiralty law — and Charles Alan Wright, co-author of the definitive treatise Federal Practice and Procedure and widely regarded as one of the most influential federal courts scholars of the twentieth century. Brian has been involved in technology, including artificial intelligence, since the early 1990s. His representative matters include the Maersk Alabama crew, Captain Wren Thomas, a $10 million-plus recovery for a boat captain who suffered a severe infection on the job, and a record-breaking $22 million-plus trucking settlement. He is the author of published books and articles on law, leadership, and technology, and hosts the Lessons from Leaders podcast.