The AI Law Firm That Reviewed $1 Billion in Contracts With 45 Attorneys

Between June 2025 and April 2026, one AI-native law firm reviewed $1 billion worth of contracts. It employed 45 licensed attorneys to do it. That firm is now Series B-funded, backed by Sequoia and Bain Capital, and explicitly targeting the work that sustains most small transactional practices.

This is not a fundraising announcement. It is an operational report on how AI-first legal competition actually works at scale — and what it means if your firm handles commercial contracts.

What Crosby Actually Does

Crosby focuses on high-volume commercial contract work: NDAs, master service agreements, data processing agreements, partnership agreements. The contracts that legal teams at technology companies, professional services firms, and mid-market businesses generate constantly and that form the steady-revenue backbone of many small transactional practices.

The firm has approximately 100 employees total: 45-50 licensed attorneys and 40-50 engineers and operations staff. Tasks that previously took hours now take 10-20 minutes per contract. The firm uses fixed-fee pricing rather than hourly billing.

In a June 2026 interview with Artificial Lawyer, co-founder Ryan Daniels described the growth: the firm went from $30 million in contract volume at seed funding in June 2025 to $1 billion in contract volume by its Series B in April 2026. That is 33x growth in under 12 months.

How the AI Works — and Why It Is Different

The standard assumption about AI at law firms is that AI drafts, a human reviews, and the human is responsible. Crosby's model adds a layer that most firms have not built: the AI agent is trained to replicate each individual attorney's specific style and judgment.

Daniels described these as "agents that sort of mimic me and my style — it becomes really scalable." The goal is not a generic contract AI but an AI that answers the same question the way that specific attorney would answer it, with that attorney's approach to risk tolerance, preferred language, and client-specific knowledge built in.

This matters because it changes the economics. If the AI outputs work that reliably reflects a specific attorney's judgment, one attorney can oversee 10x or 20x the contract volume they could previously handle. The licensed attorney is still accountable. But the per-contract economics shift dramatically.

The firm also built its own benchmark — Crosby Intelligence — to measure performance on multi-step contract negotiations, which suggests they are already thinking about quality measurement in a systematic way that most firms have not formalized.

The MSO Structure — And Why It Keeps Appearing

Crosby operates through an MSO structure: two legal entities, one a licensed law firm (which can only be owned by attorneys under most state rules), the other a separate technology and operations corporation that can accept outside investment.

This structure allows Sequoia and Bain Capital to invest in the technology company without technically owning the law firm itself — which would violate bar regulations in most US jurisdictions.

What makes this structurally significant is that it is the same vehicle private equity firms are using to acquire law firms from the opposite direction. As covered in the wave of PE law firm MSO acquisitions, PE funds are setting up MSOs to take indirect stakes in law firm economics. Holland & Knight has led 17 of these deals in 2026 alone.

Two different types of capital — venture capital building AI-native firms from scratch, and private equity acquiring established firms — are using identical legal structures to enter the same legal market simultaneously.

What Daniels Says About Traditional Law Firm Competition

Daniels made an explicit competitive argument in the June 2026 interview. He said that Big Law faces structural barriers to AI adoption because of billable-hour economics: when revenue is tied to hours worked, there is no incentive to reduce the hours required per matter.

His framing positions AI-native "NewMod" firms (new model law firms) as likely disruptors once they move beyond high-volume commercial contracts into more complex advisory and litigation work.

That migration has not happened yet at scale. Crosby's current focus is the commercial contract category — NDAs, MSAs, DPAs. Complex litigation, regulatory work, and specialized advisory remain territory where AI-native firms have not yet demonstrated the same speed and volume economics.

The honest assessment: Crosby is currently a competitive threat for the transactional work, not for the full practice scope of a general-service small firm. But the direction is clear.

What This Means for Your Firm Right Now

The Crosby data point is useful because it is specific enough to act on.

First: identify your exposure. What percentage of your revenue comes from high-volume commercial contracts — NDAs, MSAs, vendor agreements, DPAs for clients concerned about data handling? If that work represents 30% or more of your revenue, the competitive pressure from AI-native firms is no longer hypothetical.

Second: understand the pricing signal. Crosby uses fixed-fee pricing. The commercial contract category is shifting from hourly to fixed-fee even outside AI-native firms. If you are still billing hourly for routine commercial contract review, you are already behind the pricing model your clients are beginning to expect.

Third: consider what Crosby cannot replicate yet. The 45-attorney team at Crosby is doing contracts efficiently because those contracts follow patterns. Your firm's value in complex or novel legal situations — the deal that has never been done exactly this way before, the dispute with regulatory dimensions, the client relationship built over years of nuanced advice — is not what they are optimizing for. Making that distinction explicit to clients is not a defensive move; it is an accurate description of where specialized judgment still sits with a human attorney.

The AI law firm that reviewed $1 billion in contracts in 12 months is not the end of this story. It is a data point that tells you where the floor is — and how fast it is rising.

---

Thomson Reuters CoCounsel and DeepJudge released an integration in June 2026 that allows law firms using CoCounsel to access their own institutional knowledge (past work product, precedents) alongside Westlaw and Practical Law in one workflow. For firms watching the legal AI infrastructure develop, that integration is relevant context alongside what Harvey is building with firm-specific model training.