I help small law firms automate workflows using the AI tools they already pay for | Ex-Big Law partner turned peer-credible translator | No hype. No tech-bro jargon.
$117 for a contract review. Attorney included.
The Small Firm BriefThe AI briefing for lawyers who didn’t go to law school to become AI people. THIS WEEKAn AI entity was just licensed to practice law and is charging $117 a contract. Kirkland & Ellis bet $500 million that the technology gap between Big Law and everyone else is going to be permanent. And the cheap AI pricing you’re currently relying on is, quietly, beginning to end. Three stories. One direction of travel. WHAT’S WORTH YOUR ATTENTIONAn AI firm just got authorized to practice law — and it’s charging $117 a contract Your client’s general counsel just found a new option. Superlegal launched this week as the first AI-powered entity in the United States formally authorized to practice law. Not software. Not a chatbot dressed up in a suit. An actual law firm. Licensed under Utah’s Legal Services Innovation Sandbox that reviews and redlines commercial contracts in under 24 hours, with a licensed attorney signing off on every review, for $117 per contract. They’re starting in construction. Subcontractor agreements, vendor contracts, commercial work. The kind of mid-market transactional volume that fills the calendar for a lot of small Florida firms between the bigger cases. The math is stark: Superlegal claims a 90% reduction in legal review costs versus traditional firms. The company was founded in 2023, has been operating in Utah under regulatory protection, and just made its first major market push. Before you file this under “Utah problem”: the sandbox model is specifically designed to test alternatives that then expand. Florida has its own regulatory sandbox for legal services. The Board of Governors created it in 2021. The pressure to expand these models nationally is only going in one direction. This doesn’t touch litigation. It doesn’t touch the relationships that keep clients coming back. But transactional work? The contract review someone calls you for between cases? That’s where the pressure builds first. Know what’s coming. The AI tools you’re paying $20 a month for are priced below cost. That’s starting to change The subscription on your AI tools is not the real cost. OpenAI, Anthropic, Google. they have all been pricing below cost to build market share. Anthropic ended subsidized pricing for enterprise accounts in April. GitHub moved its AI coding tool to usage-based billing on June 1. Law.com ran a piece last month that almost no small-firm newsletter touched: “The Era of Subsidized AI Tokens Is Ending. Where Does That Leave Law Firms?” The Claude Pro subscription at $20/month and ChatGPT Plus at $25/month are still flat-rate and still below cost. That won’t last indefinitely. Here’s the read most newsletters won’t give you: this makes right now the best time to build your AI habits. Not because of the technology. Because of the economics. Every workflow you build today, intake drafts, billing narratives, client summaries, locks in value before prices move. The muscle memory you build now is worth twice as much when subscriptions increase. The firms that haven’t started yet will pay more to learn the same skills later. The urgency argument for AI adoption always used to be “you’re falling behind.” Now it has a dollar sign attached too. Kirkland & Ellis committed $500 million to AI they will never share with you This story got picked up in Big Law circles and almost nowhere in small-firm newsletters. That’s backwards. In May, Kirkland & Ellis, the highest-grossing law firm in the country, announced a $500 million investment over the next three to four years to build proprietary AI tools. Not licensing Harvey. Not deploying Copilot. Building their own, with 180 technologists and 250 Kirkland lawyers as design partners. The finished product will be built around Kirkland’s institutional knowledge, preferred workflows, and client requirements. And it will never be sold to another firm. This is not a story about Kirkland being rich. It’s a story about the technology divide in this industry becoming structural. Here’s the part that matters for your firm: Kirkland’s tools won’t be ready for years. What’s available off the shelf right now — Claude, ChatGPT, Clio Duo, CoCounsel, Descrybe — is already capable. The gap between a Kirkland partner and a sharp small-firm lawyer who has deployed AI well is smaller today than it will be in 2029. That window closes when Kirkland’s platform goes live. It’s open right now. TOOL OR WORKFLOW OF THE WEEKDescrybe — Free legal research with a built-in citator Descrybe (descrybe.ai) is a legal research platform with access to 3.6 million AI-summarized judicial opinions, natural-language search, and a built-in citator to check whether a case is still good law. In March 2026, they launched DescrybeLM — their own legal reasoning engine built specifically for case research, not adapted from a general-purpose model. The free tier requires no login and no email address. Open it, search, read. For solo practitioners doing routine research, it works from the first click. What separates Descrybe from asking the same question in ChatGPT: your queries don’t leave the platform. Descrybe doesn’t route your search through OpenAI or Anthropic. For a Florida attorney conscious of client confidentiality obligations under Ethics Opinion 24-1, that distinction is real. You’re not feeding a client’s fact pattern into a general-purpose AI’s training pipeline. Cost: Free (no login required) / $20/month for commercial use One caveat: Descrybe’s coverage skews toward federal and major state courts. Before relying on it for Florida-specific research, verify your jurisdiction’s caselaw is adequately indexed. For anything high-stakes, cross-check against Westlaw or Lexis. PROMPT OF THE WEEKBilling Narrative Generator Every lawyer’s time entries look the same: “Researched issue, 2.4 hrs.” “Call with client, 0.7 hrs.” “Reviewed documents, 1.8 hrs.” These entries are defensible on an audit and completely opaque to the client reading the invoice. Your billing software captures the time. It doesn’t narrate the work. That gap is yours to close and AI closes it fast. Paste this into Claude or ChatGPT with your actual time entry notes substituted in the brackets: PROMPT You are a billing narrative writer for a small law firm. Convert each time entry below into a polished billing narrative for the client invoice. Each narrative should:
Be specific about what was done (no vague language like “researched issue” or “reviewed documents”)
Lead with the task, not the time
Use complete sentences, professional tone
Sound like the work was done by an expert, not logged by a timekeeper
Be 1–3 sentences maximum per entry
Do not add or invent details not present in the notes. If an entry is too vague to narrate specifically, flag it with [NEEDS CLARIFICATION] rather than guessing.
TIME ENTRIES: [Paste your raw time entries here, one per line, with date, description, and hours]
ONE MORE THINGSmall firms have every structural advantage for adopting AI fast. No IT approval process. No partner committee. No data governance council deciding whether Claude is on the approved vendor list. A managing partner at a five-person firm in Gainesville can decide to start using an AI tool today and be using it tomorrow. One conversation, done. And yet, consistently, the data shows solo and small-firm adoption lagging behind large firms. The lawyers with the least institutional friction are moving the slowest. This is not a technology problem. It’s a permission problem. Small-firm lawyers are waiting for someone to tell them it’s safe, a definitive bar opinion, a malpractice carrier statement, a colleague who went first. The same instinct that makes lawyers excellent at managing client risk is making them slow to manage their own. Nobody is coming to issue that permission slip. The Florida Bar published Ethics Opinion 24-1 eighteen months ago. The tools are available. The economics are still favorable, and that is going to change. The firms that haven’t started are not being careful. They’re running out of time to be early. The only thing standing between a 10-person Florida firm and a meaningful AI advantage is the decision to go first. — Greg Actionable AI | Less hype. More results.' Greg Yates · Actionable AI for Lawyers · actionableai.net · gyates@actionableai.net · (212) 765-0685 You’re receiving this because you signed up at actionableai.net. No hard feelings if it’s not for you. Unsubscribe below. Sources: - Superlegal Launches the AI Law Firm for U.S. Construction — GlobeNewswire, June 3, 2026 - The Era of Subsidized AI Tokens Is Ending. Where Does That Leave Law Firms? — Law.com, May 26, 2026 - Legal AI Has A Growing Token Price Problem — Artificial Lawyer, June 3, 2026 - Token Costs and the Future of Law Firm AI Spend — Artificial Lawyer, June 4, 2026 - Kirkland & Ellis Investing $500 Million to Build AI Platform — Bloomberg Law, May 2026 - Free Legal Research Site Descrybe.ai Launches A Paid Suite of Legal Research Tools — Lawnext, June 2025 Greg Yates June 10, 2026~6 min read |
Actionable AI for Law
I help small law firms automate workflows using the AI tools they already pay for | Ex-Big Law partner turned peer-credible translator | No hype. No tech-bro jargon.
The Small Firm Brief The AI briefing for lawyers who didn't go to law school to become AI people. Issue #3 · June 3, 2026 This week. Three stories, one theme: the AI policy is not enough anymore. The question courts, clients, carriers, and opposing counsel are going to ask is not whether your firm allows AI. It is whether you can prove who used it, what they used it for, and what a lawyer verified before anything left the firm. What's worth your attention. 1. Your AI policy is not a defense....