For most of the twentieth century, drafting a will was a reliable profit center for small-firm attorneys. The documents were formulaic enough to assemble from templates, yet sufficiently intimidating to justify hourly rates that could stretch a simple estate plan into four figures. The arrival of consumer software like LegalZoom in the early 2000s began chipping at this model. But it was the integration of large language models into legal workflows that finally cracked it open.
The shift happened not with a bang but with a quiet repricing. When AI tools became capable of generating coherent, jurisdiction-appropriate testamentary documents from a few intake questions, the economics of will-drafting collapsed for attorneys who had been selling document production disguised as legal counsel. A task that once took three billable hours could be accomplished in minutes. The commodity was exposed as a commodity.
The bifurcation
What emerged was a two-tier market that had always existed in latent form. On one level, consumers with straightforward estates—single property, no business interests, uncomplicated family structures—found they could obtain adequate documents through AI-assisted platforms for a fraction of traditional fees. These tools do not practice law, but they produce drafts that competent users can review and execute.
On the other level, estate planning attorneys who had always emphasized counseling over drafting discovered their practices were insulated, even enhanced. The complex work—blended families, business succession, charitable planning, tax optimization across generations—requires judgment that no language model can replicate. These practitioners found that AI had eliminated their least interesting work while clarifying their value proposition to clients who actually needed them.
What the machines cannot do
The limitations are instructive. A language model can generate a pour-over will that coordinates with a revocable trust. It cannot sit across from a widow and discern that her hesitation about naming her son as executor stems from his undisclosed gambling problem. It cannot navigate the conversation where a business owner realizes his succession plan will destroy his partnership. It cannot recognize when a client's stated wishes mask family dynamics that will guarantee litigation.
Estate planning, at its core, is about mortality and money—the two subjects people lie about most readily. The attorney's role has always been part therapist, part translator, part referee. AI automated the translation. The rest remains stubbornly human.
The access question
Critics of the legal profession have long argued that routine legal services were artificially expensive, leaving middle-class Americans without adequate estate planning. By some estimates, more than half of American adults lack even a basic will. If AI-assisted tools bring that number down, the social benefit is real, whatever the cost to attorneys who built practices on document assembly.
The counterargument is that cheap documents are not the same as good planning, and that consumers may not know the difference until probate reveals the gaps. A will that fails to account for state-specific formalities, or that creates unintended tax consequences, is worse than no will at all. The democratization of legal documents without the democratization of legal judgment is an incomplete revolution.
Our take
The estate planning bar spent decades conflating two distinct services: producing documents and providing counsel. AI called the bluff. The attorneys who survive and thrive will be those who always understood the difference—who saw the will as a byproduct of the planning conversation, not its purpose. For everyone else, the adjustment will be painful but clarifying. The profession is not dying. It is finally being forced to articulate what it does.
