Here is something BigLaw partners do not want you to know: at least 42% of the Am Law 100 are already using AI to do first-pass contract review. They will not say which AI. They definitely will not tell clients they are billing $800 an hour for work that an algorithm did in 12 minutes.
The Quiet Revolution Nobody Will Admit To
I have spent the past year talking to associates, partners, legal ops directors, and in-house counsel at companies ranging from Series B startups to Fortune 100 enterprises. The conversations almost always follow the same script. Off the record, they describe AI tools that have fundamentally changed how they review contracts, conduct due diligence, and prepare for negotiations. On the record, they talk about “exploring innovative solutions” and “maintaining the highest standards of professional rigor.” The gap between those two statements is where the real story lives.
Let me give you a number that should make every client paying hourly legal fees sit up straight. In August 2025, Harvey AI announced it had crossed $100 million in annual recurring revenue and was being used by 42% of the Am Law 100. That is not a pilot program. That is not a handful of tech-forward firms dipping their toes in. That is nearly half of the most powerful law firms on the planet paying for a subscription to an AI that reads contracts, drafts memos, and summarizes case law. And yet, walk into most of those firms and ask a partner directly whether they use AI, and you will get a carefully worded non-answer about “evaluating tools.”
The associates know, though. They always know. And they are caught between two fears: that the AI will take their jobs, and that the partners will find out they are already using it on the side to keep up with impossible workloads.
What These Tools Actually Do (And How Well They Do It)
Harvey AI is the one that gets the most attention, partly because of its OpenAI backing and partly because of its relationship with A&O Shearman, the global law firm formerly known as Allen & Overy. A&O started trialing Harvey in late 2022, and by early 2023, roughly 3,500 of its lawyers across 43 offices had submitted around 40,000 queries for live client work. That was the trial. By 2025, the partnership had evolved into something far more ambitious: the two organizations are now building agentic AI agents together — multi-step reasoning systems that can handle complex legal workflows for antitrust filing analysis, cybersecurity assessments, fund formation, and loan review. Harvey raised its valuation past $8 billion in late 2025. This is not a side project anymore.
CoCounsel is Thomson Reuters’ answer, built on the bones of Casetext, which Thomson acquired and folded into its empire. The product is now embedded across Westlaw and the broader Thomson ecosystem, which means it has distribution that startups can only dream of. CoCounsel analyzes contracts, extracts terms, dates, and dollar amounts, and identifies compliance issues and risk provisions. A million professionals are now using it, according to Thomson Reuters’ own March 2026 numbers. The pricing is enterprise-grade — CoCounsel Core starts at $225 per user per month — but for firms already paying for Westlaw, it integrates seamlessly. The newest iteration, CoCounsel Legal, introduces guided multi-step workflows: upload a contract, and the system walks you through review, risk flagging, and suggested redlines in a single conversational thread.
Luminance is the specialist that terrifies M&A lawyers the most, because it does exactly what those lawyers spent their first five years of practice doing: reading rooms full of documents and flagging anomalies. Luminance uses machine learning to read contracts at scale, automatically identifying key clauses, deviations from standard terms, and risk indicators. Slaughter and May — one of the Magic Circle firms — used Luminance to cut their due diligence process by 75% and shorten M&A deal timelines by 30%. The platform claims contract review time savings of up to 90% and a reduction in external counsel spend of up to 90%. Even if you halve those numbers for marketing inflation, the implications are staggering.
ContractPodAi (rebranded as Leah) takes a different approach. Instead of being a tool lawyers use, it is an agentic operating system that enterprises deploy across legal, procurement, finance, and HR. Leah generates contracts from guided inputs, flags risk language, suggests revisions consistent with internal playbooks, and tracks obligations after execution. It was named a Leader in the 2025 IDC MarketScape for AI-Enabled CLM and placed in the 2025 Gartner Magic Quadrant. Its newest trick is the Leah Tariff Agent, which automatically identifies tariff-related provisions in existing contracts — the kind of capability that became urgently relevant when trade policies shifted in early 2026.
| Platform | Best For | Key Stat | Starting Price |
|---|---|---|---|
| Harvey AI | Multi-practice BigLaw | 42% of Am Law 100 | Enterprise (custom) |
| CoCounsel | Westlaw-integrated review | 1M+ professionals | $225/user/mo |
| Luminance | M&A due diligence | 75% faster diligence | Enterprise (custom) |
| Leah (ContractPodAi) | Enterprise CLM | Gartner MQ placement | Enterprise (custom) |
| Spellbook | Solo / small firm drafting | Word-native interface | $99/user/mo |
The Billing Problem Nobody Wants to Talk About
Here is where the conversation gets uncomfortable. The traditional law firm model bills by the hour. A first-year associate at a large firm bills somewhere between $400 and $700 per hour. A significant chunk of that associate’s time — 40 to 60 percent, depending on the practice area — goes to document review, contract analysis, and legal research. These are precisely the tasks where AI delivers its most dramatic time savings.
So what happens when a task that used to take a junior associate 15 billable hours can now be completed in 45 minutes? Mathematically, the firm has two options. Option one: pass the savings to the client, billing for the actual time spent. Option two: bill something closer to the old rate, pocketing the difference. In practice, most firms are doing something messier — a blend of reduced hours with vague “technology surcharges” or “project management fees” that partially offset the efficiency gains.
Florida Opinion 24-1 now mandates that attorneys disclose AI use when it impacts client billing or costs. Other state bars are actively debating similar requirements. The American Bar Association has flagged AI disclosure as a professional responsibility issue. And yet, according to attorneys I spoke with at five different firms, the internal guidance is still largely “do not volunteer the information unless asked.” One litigation associate at a mid-size firm told me, bluntly: “We use Harvey for first-pass memo drafts. It saves me three hours minimum. I bill for two. My supervising partner knows and does not care because the client does not ask.”
This is not a scandal. It is a structural tension baked into the hourly billing model that the profession has resisted changing for decades. AI did not create the problem. It just made it impossible to ignore.
What This Means If You Are Not a Lawyer
If you are a business owner, in-house counsel, or anyone who regularly signs contracts, the adoption of AI legal tech has practical implications for how you should think about your legal spending. And honestly, some of this is good news.
First, the quality floor has risen. AI contract review tools do not get tired at 2 AM. They do not skip page 847 of a merger agreement because they have been reading for nine hours straight. They do not miss a non-standard indemnification clause buried in an exhibit because their attention drifted. Stanford research found error rates of 17% for Lexis+ AI and 34% for Westlaw AI-Assisted Research — which sounds alarming until you compare it to the error rates of exhausted junior associates on their third consecutive all-nighter during a deal sprint. AI is not perfect. It is consistently imperfect in predictable ways, which, paradoxically, makes it easier to quality-check than a human who makes random, fatigue-driven mistakes.
Second, you can now demand transparency. Ask your outside counsel directly: “Are you using AI tools for contract review or document analysis on our matters?” If they say no, they are either behind the curve or not being honest. If they say yes, follow up: “How does that affect your billing?” You have a right to know. Florida has already codified this. Other jurisdictions will follow.
Third, some of these tools are available to you directly. You do not need a law firm to run a contract through Luminance or feed an NDA into CoCounsel. These platforms are increasingly selling to corporate legal departments and even mid-market businesses. The question is not whether AI will review your contracts. It is whether you are paying a law firm to be the middleman in that process, or whether you are doing it yourself and only escalating the genuinely complex issues to human lawyers.
Frequently Asked Questions
Yes, and you should not feel awkward about it. AI-assisted contract review is not inherently bad — in many cases it produces more thorough results than purely manual review. But you deserve to know how the work is being done and how it affects what you are paying. A good lawyer will answer this question directly and explain how AI fits into their quality control process. A lawyer who gets defensive about the question is telling you something important.
For routine contracts like NDAs, standard vendor agreements, and employment offer letters, yes. Tools like Spellbook, CoCounsel, and Luminance can flag unusual clauses, missing provisions, and deviations from market-standard terms. But for anything high-stakes — M&A documents, complex licensing deals, IP assignments with significant value — AI should supplement legal counsel, not replace it. The AI can tell you what the contract says. It cannot tell you whether what it says is actually in your best interest given your specific business situation.
Ask about their verification process. Reputable firms using AI have a human-in-the-loop workflow: the AI does the initial analysis, and a qualified attorney reviews every output before it reaches you. Red flags include receiving unusually fast turnaround on document-heavy projects without any discussion of methodology, or finding generic analysis that does not reflect the specific nuances of your deal. If something feels templated rather than tailored, ask questions. The best firms are transparent about their process and proud of how they integrate AI with human expertise.