A few years back I was closing a business deal and needed a contract reviewed. My attorney was thorough, professional, and billed me for four hours of work — most of which, I later realized, was reading and flagging standard clauses I could have identified with a basic checklist. The advice on the two non-standard clauses? Genuinely valuable. The document review portion? I was paying $400 an hour for something that today's AI does in about 30 seconds.
I'm not blaming the attorney. That's just how the work has always been done. But "how the work has always been done" is exactly the sentence that should make any professional nervous right now.
Law is one of the most AI-disrupted professions in the world right now — and most people outside the industry have no idea how fast it's moving. This is series entry #2 in No Industry Is Safe. Last time we talked about healthcare. This time: law.
What's Already Happening
Legal AI has moved from novelty to mainstream faster than almost any other professional service. Major firms are already using it at scale. The question isn't whether it's coming — it's whether your firm is going to lead the adoption or get undercut by the ones that do.
-
Contract review — the first and biggest hit Tools like Kira Systems, Harvey, and ContractPodAi can review thousands of pages of contracts in minutes, flag non-standard clauses, identify missing provisions, and summarize risk exposure. Work that used to take junior associates days now takes AI minutes — with accuracy that rivals (and sometimes beats) a tired associate at 11pm billing their sixth hour.
-
Legal research — Westlaw and Lexis are already AI-first Both Westlaw and LexisNexis have embedded AI research assistants that don't just search case law — they synthesize it. They find the relevant precedents, summarize the arguments, and surface the cases your opposition is likely to cite. What used to be a first-year associate's bread-and-butter work is now an AI query with a follow-up prompt.
-
E-discovery — AI was born for this Document review in litigation used to mean armies of contract attorneys reviewing millions of documents for relevance and privilege. AI e-discovery tools now process those same documents in a fraction of the time, with comparable accuracy, for a fraction of the cost. This is one of the largest line items in complex litigation — and it's being compressed aggressively.
-
Document drafting — AI is a surprisingly good first draft NDAs, employment agreements, LLC operating agreements, IP assignments — AI tools can draft competent first versions of standard legal documents in seconds. Harvey (built on GPT-4 and fine-tuned on legal corpora) is being used by Allen & Overy, one of the world's largest law firms, for exactly this purpose. The attorney reviews and refines. The AI does the grunt drafting.
-
Client intake and triage — AI at the front desk AI chatbots are now handling initial client intake — gathering facts, assessing case viability, flagging jurisdictional issues, and routing clients to the right practice area. For solo practitioners and small firms, this is the equivalent of a full-time paralegal working 24/7 without benefits.
What AI Can't Replace
Here's where I'll give the lawyers their due — and it matters, because there's genuine value that AI can't replicate yet.
Judgment in ambiguous situations. The law isn't just rules — it's interpretation, strategy, and prediction. A seasoned litigator reading a judge's tendencies, anticipating how opposing counsel will move, and advising on whether to settle or fight — that's pattern recognition of a different kind. Deeply contextual, built from years in rooms that AI hasn't been in.
Negotiation and relationships. Deal-making in M&A, partnerships, and complex commercial agreements isn't just about the contract language. It's about the relationship, the leverage, the read on the other party's actual priorities. An AI can draft the terms. It can't read the room at the closing table.
Courtroom presence. Cross-examination, oral argument, jury persuasion — these are performance skills. The ability to adapt in real-time to a hostile witness or a skeptical judge is not a text prediction problem. Not yet.
Accountability and privilege. Attorney-client privilege, malpractice liability, bar licensure — the legal framework around legal services is built around human professionals. Until that changes structurally, there's a floor under the human lawyer's role that AI can't penetrate on its own.
The Real Disruption Nobody's Talking About
The real disruption isn't "AI replaces lawyers." It's "one attorney with AI does the work of five attorneys." Which means firms run leaner. Junior associate pipelines shrink. And small firms that don't adopt are competing against AI-augmented firms that can do the same work cheaper and faster.
The clients — especially small and mid-size businesses — are going to figure this out. They're already starting to. When a business owner can use a tool like DoNotPay or Lawfully to handle routine legal tasks for $50/month instead of $400/hour, they will.
The deeper issue is the billable hour model itself. That model was built on information asymmetry — clients paying for access to expertise they couldn't easily get elsewhere. AI is systematically eliminating that asymmetry. Research, document review, drafting, intake — all of it is becoming more accessible and less exclusive.
What survives? The high-stakes, high-judgment, high-relationship work. Complex litigation. M&A advisory. Criminal defense at trial. Tax strategy for complex estates. The work where the stakes are too high to trust a first draft and the judgment required is genuinely irreplaceable.
The solo practitioner doing boilerplate wills, standard business formation, and routine contract reviews is in the most exposed position. That work is being commoditized — not by other lawyers, but by software.
What It Means If You Run a Law Firm
Adopt AI tooling for document work immediately. Contract review, research, and first-draft drafting should already be AI-assisted. If your associates are doing this work manually, you're paying for something that's being automated. That's money leaving your firm and going into your competitors' margins.
Reposition around judgment, not process. The firms that will win are the ones that market themselves on the quality of their strategic advice, not the quantity of documents they can process. If your pitch to clients is "we're thorough" — that's an AI pitch now. If it's "we have the best judgment in the room on your specific kind of problem" — that still holds.
Use AI to expand capacity, not just cut cost. An attorney who uses AI assistance can handle more clients, respond faster, and offer more competitive pricing without sacrificing margin. That's a competitive moat — for now. In two years, it's table stakes.
The lawyer isn't going away. The $400/hour document reviewer is.
The legal profession isn't getting replaced — it's getting restructured. The high-judgment, high-stakes, high-relationship work is safe and may actually become more valuable as AI handles the commodity layer. Everything else is under pressure right now, not eventually.
The attorneys who treat AI as a productivity multiplier will look back on this moment as an advantage. The ones who resist it will spend the next five years wondering why their margins keep shrinking.
Next in the series: Finance & Accounting — your accountant is about to have a lot more free time.