AI As Law Firm—or Not: A U.S. Perspective on Access and Regulation
INTERVIEW ON THE PRICE OF BUSINESS SHOW, MEDIA PARTNER OF THIS SITE.
Recently Kevin Price, Host of the nationally syndicated Price of Business Show, interviewed Alexander Paykin.
The Alexander Paykin Commentaries
AI as Law Firm—or Not: A U.S. Perspective on Access and Regulation
In the United Kingdom, Garfield AI has made headlines as the first fully AI-powered “law firm” to win regulatory approval from the Solicitors Regulation Authority. By offering services such as £2 debt-collection letters, Garfield has demonstrated that artificial intelligence can deliver legal help at scale and at an unprecedentedly low cost. For the UK, this is a bold regulatory experiment in broadening access to justice.
From a U.S. perspective, Garfield’s emergence raises an uncomfortable question: could anything like this happen here?
The short answer is: not under the current rules. In most American states, nonlawyer ownership of law firms is prohibited under professional responsibility regulations. An entity like Garfield, with no traditional lawyers at the helm, would likely be considered unauthorized practice of law. This restriction reflects a long-standing concern with protecting the independence of legal judgment and ensuring accountability, but it also creates barriers to innovation.
At the same time, the access-to-justice crisis in the United States is more acute than in many other developed countries. Millions of Americans face legal issues each year—debt collection, landlord-tenant disputes, family matters—without legal representation, largely because of cost. A system like Garfield, offering ultra-low-cost standardized services, could address gaps that the traditional legal market has failed to fill. That makes the contrast with the U.S. all the more striking.
Some incremental change is occurring. Arizona and Utah have launched regulatory “sandboxes” that permit alternative business structures and technology-driven legal service providers to operate under supervision. These experiments are still limited, but they suggest a pathway where AI-enabled legal services could gain traction in the U.S. if regulators are willing to balance innovation with oversight. Whether these state-level efforts expand—or stall—will be a key determinant of whether an “AI law firm” model can exist here.
Meanwhile, at the top end of the U.S. market, the story resembles that of the UK’s Allen & Overy. Major firms are not seeking to replace lawyers with AI, but rather to integrate AI into their operations to improve speed and efficiency. Platforms like Harvey, already in use at global firms, are being tested across U.S. practices for tasks such as contract review, research, and compliance support. The American approach, at least for now, is one of augmentation, not substitution.
This divide highlights two futures playing out in parallel. One is AI as a tool to democratize access to justice, filling gaps in routine, high-volume matters where clients cannot afford lawyers. The other is AI as an amplifier of elite expertise, enabling top firms to serve corporate clients faster and more strategically. In the U.S., the second model is flourishing, while the first remains largely blocked by regulation.
The Garfield experiment in the UK underscores what is possible when regulators allow space for new models. For the U.S., the challenge is sharper: can the legal system reconcile its restrictive rules with the urgent need for affordable legal help? Or will AI’s benefits remain confined to those who can already afford the best legal services?
What is clear is that AI will play an increasingly central role in law, whether as a firm itself or as a powerful assistant within one. The question for the U.S. is not whether AI will transform the market, but whether regulators will allow it to address the access-to-justice crisis—or whether innovation will remain the privilege of the already well-served.
Check out more national news stories here.
Explore more insights at https://thedailyblaze.com/.
About Author
