What AI Can't Commoditize | Relevant
Software can now draft the contract. It still can't be the person a client trusts with a life-changing decision.
Published 2026-05-27 — Updated 2026-06-21
Ask an artificial-intelligence model to draft a will, a lease, or a shareholder agreement, and it will hand you something serviceable in seconds. Ask it to sit across a table from a business owner the week a deal is falling apart, read what isn't being said, and counsel a decision that will shape the next ten years of that person's life — and it has nothing to offer. The distance between those two requests is the whole future of the legal profession.
For two centuries, lawyers sold two things bundled together: the document and the judgment. The document was the visible product — the contract, the filing, the memo. The judgment was the thing underneath it — knowing which document, and why, and when, and what it meant for this particular client in this particular moment. For most of history the two were inseparable, and clients paid for the bundle by the hour.
Artificial intelligence has pulled the bundle apart. The document — the drafting, the research, the first pass — is being commoditized in front of us, and not slowly. What's left, suddenly exposed, is the judgment. And judgment turns out to be stubbornly resistant to automation. As one MIT economist observed, the law's low tolerance for error and the genuine complexity of real problems put hard limits on what a model can responsibly do. A machine can tell you what the rule is. It cannot tell you whether to fight or to settle, whether the person across the table can be trusted, or whether the technically correct answer is the wise one. It cannot be accountable. It cannot be trusted, the way a person is trusted, with the things that matter most.
This is why the value in law is moving up, not away. The work AI commoditizes was never where the deepest value lived; it was only the most visible part of it. What clients have always actually paid for — whether they could name it or not — is the feeling of being in good hands: that a real person who understands their situation is watching out for them, and that when the stakes are high and the path is unclear, someone with experience and judgment is on their side and will own the outcome. No model provides that. It is the most human part of the work, and it is the part that endures.
Some of the field's sharper analysts now argue that AI may even increase the demand for that kind of counsel. As routine legal help becomes cheap and abundant, people run into more moments where they realize they need a human they can trust, not just an answer. Abundance at the bottom of the market tends to raise the value of judgment at the top.
For a lawyer, that is clarifying rather than frightening. The goal is not to compete with the machine on the work the machine does well. It is to let the machine do that work — quickly, cheaply, tirelessly — and to spend your own scarce hours where you are irreplaceable: in the room, in the relationship, in the judgment. That means building a practice organized around trust rather than transactions, around being someone's lawyer rather than producing someone's document.
That orientation is exactly what a modern platform is built to make possible. The technology handles what should be automated. The brand earns the trust that brings a client through the door. The infrastructure carries the business, so the attorney's attention is free for the part that counts. The firms that win the next decade will not be the ones that produced the most documents. They will be the ones a client could not imagine facing a hard decision without.