WC.com

Tuesday, November 17, 2026

Where Do We Go from Here?

Horace Middlemier* called me recently to discuss several topics, including Management Services Organizations (MSOs) and Alternative Business Structures. The crux of their perceived value is that only lawyers can own law firms in most states. That means that the venture capitalists may struggle to use their funds to leverage technology or efficiency in the legal practice.

Venture capitalists invest in new businesses and receive equity in them. If the business is already up and running, similar investors may provide Private Equity. Their infusion may simply fuel expansion of an existing idea, product, or service. Or, with their ownership may also come their influence and ideology and a shift in the company's business, strategy, or ideals. Some see its fingerprints on Red Lobster, "Sears, Mervyn’s and ShopKo, ... Steward Healthcare and Manor Care."

Some see this as an opportunity for better corporate funding and management. Others see this as a threat. There has been significant venture capital investment in the medical markets. Investors have acquired hospitals, practices, and other facilities. There are critics. 

Now some want to make similar investments in the legal practice market. They are somewhat frustrated by the state's law firm ownership rules, leading to discussion of the MSO solution. This is "a separate business entity—typically owned by nonlawyers—that provides operational, administrative, and business services to a law firm." The pitch is that the business side is handled by the experts, leaving the lawyer to focus on the law.

Such an arrangement allows the law firm to raise money beyond the typical debt option. Any company, or law firm, can borrow money and then work to service the debt. But in the broadest sense, most companies also have the alternate path of equity—selling stock. In the law firm world, you might argue that a similar equity path is available through adding partners. Nonetheless, the partners have to be lawyers, thus limiting the supply of potential investors. 

Why the sudden capital demand in legal practice? One firm notes that
"Modern legal practices demand sophisticated business operations, including data analytics, cybersecurity, marketing automation and advanced HR systems."
Tech is attractive. IT can be leveraged and may convey advantages in efficacy or efficiency or both. The MSO provides such services and perhaps more. The venture or private capital invests in the tools and then, essentially, leases those tools to the lawyers. Perhaps more? How about the promised land of artificial intelligence (AI)? AI is drawing huge investment. Its promise is directly tied to efficacy and efficiency.

The MSO concept allows rapid law firm deployment of "Advanced AI tooling, data integration, and custom agent workflows (that) require massive capital outlays." But there is a risk.

If you have not noticed, AI tools can be unwieldy, erratic, obsequious, disobedient, pedantic, and wrong. Yes, those are some of the good qualities. I am no more advocating for or against AI, but caution and introspection should accompany any change. 

What are the solutions or guardrails for AI? One that is highly recommended is isolation. That is, using the AI Large Language Model (LLM) but restricting its resources. By now, everyone has heard a story or two about LLMs hallucinating and presenting internet drivel as if it were indisputable truth. So, don't let your LLM look at the internet. Only allow it to look at your prior work product.

That is called a "RAG," (Retrieval-Augmented Generation), and many are already using the concept. Hundreds of millions have already been invested by lawyers and developers. With such a tool or process, the lawyer could ask an LLM to generate a Motion for Summary Judgment today and to use only previous such motions that the lawyer her/himself has previously drafted/filed. The result may be somewhat an echo chamber, a recycling of one's own prior thoughts ("every problem starts to look like a nail"). 

Add to the mix the recent realization that law firms may not possess much in terms of inherent value. Yale Law School recently reported that a "prominent law firm" dissolved, virtually overnight. The firm had been in business for almost a century. It was perceived as large, stable, and valuable. It was gone very rapidly. 

Essentially, the Yale professor sees its weakness in the partnership structure. He describes partners leaving the firm and taking their book of business (clients) with them. When one prominent lawyer leaves, then perhaps another follows. As more owners see their collective value threatened, more depart and take their value elsewhere. 

He says the result "resembles a run on the bank.” As each strives to escape with as much as they can, those left behind begin to reconsider leaving themselves. The firm is not the value; the people are. When the people leave, what remains is a husk.

Why would a particular client stay with the firm when "their lawyer" leaves? Certainly, a firm might have prestige or name recognition. But, as likely, the client has familiarity and confidence with the individual lawyer(s) who provide the services and advice. Their faith and trust is in the person and will remain if that person wears a different team jersey. 

The scenario is logical. The potential for better financing and management is patent. The problem with the MSO investment versus the frailty of a lawyer's longevity is concerning. But it is likely deeper. 

Pause a moment to consider the 2023 Hollywood strike. Those creative types (writers, actors, etc.) were troubled that they might be paid to make a movie or show. The owners of that show might then use an AI LLM to "take an actor’s image and insert it into any scene" in another show or movie for which the actor earns no pay and in which she/he has not even participated.

The LLM might as easily be used to write a script or book using the "voice" of some specific writer or creator. The result is not "by Hemingway" but might easily "sound like Hemingway." That is because the LLM can study Hemingway, the producer, and counterfeit something that emulates Hemingway. Hemingway earns nothing on the knockoff. 

The LLM cannot counterfeit the listener or audience. Whether the LLM-produced script or film appeals to an audience is another thing altogether. But what if you could emulate the audience? The challenge becomes more complex.

What if you asked the LLM to draft a motion for you as described above, emulating your own work, but also asked it to consider the prior rulings of the assigned judge, Judge Beady Maxirmal**, and to tailor the filing to cater to this judge's thoughts, predictions, and prior rulings? The LLM is emulating a producer's voice and style and is focused on maximum impact on the listener/audience. 

As the lawyer performs such work using the LLM, who owns the resulting output (motion)? It may be filed in a single case, but in the process it becomes part of the database. That database might be owned by the law firm, or, more likely, the MSO built beside it by the venture or equity capitalists. Despite the potential for coincident ownership interests and interrelationship, the firm and MSO are distinct. 

Can they (the firm or MSO) sell that lawyer's efforts, by selling the database, to the next lawyer that comes down the street? Does it matter if that is a new lawyer hired by the law firm or some new subscriber doing business with the MSO?

The implications are potentially deeper still. Could a client (XYZ Corp.) insist that their outside lawyer use the client's LLM system (purportedly to preserve any potential privileges or to save expense)? As the lawyer works on the motion, for whatever the agreed pay, she/he simultaneously benefits the client (output, motion) and trains the client's LLM, enhancing their tool (input). The lawyer is training a tool to emulate their own work as much as an actor might train an LLM to imitate their expressions, inflections, and timing.

As the client's LLM gains sophistication and expertise, might it rely less on the particular lawyer? Might an assignment devolve from drafting a particular motion for summary judgment to "checking its cites and references in such a motion that XYZ Corp's LLM produced?" Even then, the lawyer would be training the client's LLM. With each iteration and interaction, the XYZ LLM would seemingly refine and improve. 

In the process, the lawyer's skill, judgment, and experience would be siphoned over time. The essence of the professional would develop and grow the LLM. In time, the lawyer would become a functionary. Her/his role could be reduced to proofreading, reading, and signing. See A Positive Step (June 2026). And, as a result, the lawyer's value would diminish. 

If all XYZ Corp. needs is a functionary to file its own LLM's motion, then the compensation for such a lawyer will persistently decrease in a competitive downcycle. If drafting such a motion used to take hours of research and contemplation, and the LLM can create it in minutes, the value of the person will likely diminish.  

The lawyer in this setting might not have the same defenses as an actor. Might someone gain traction with copyright, contract, or even legislation? Perhaps for the actor, less likely for the lawyer. But it is even less likely the judge would. 

In the worst case, the value of each role diminishes with each successive iteration. Henry's prognostication is fulfilled. Without killing a single lawyer, the LLMs and MSOs kill the legal profession. See The First Thing We Do (April 2026). I have Aretha in my head right now: "Who's Zoomin' Who" Arista, 1985). 

*Horace is not a real person. The name is a foil, an example, an "everyman," a literary device.

** Judge Maxirmal is likewise not real. The name is a foil, an example, an "everyman," a literary device.