Y Combinator alum Moritz raises $9m after law firms rejected its AI tools — so it became a law firm itself

Pamir Ehsas spent years advising OpenAI on contracts. When he tried to sell AI tools to law firms, they said no.

So he became a law firm instead.

Moritz, the company Ehsas co-founded with MIT-Fulbright engineer Stefan Mandaric, has raised $9 million in pre-seed funding — closed in four days — backed by Y Combinator, 20VC, and a roster of angel investors that reads like a who's who of tech: founders from Reddit, Dropbox, Hugging Face, Instacart, and Runway, plus employees of OpenAI and ElevenLabs. The round was originally sized at $3 million. Ehsas and Mandaric stopped taking checks after day one (Business Insider).

The funding is fast. The origin story is faster.

Ehsas and Mandaric started as Arcline, building AI software to sell to law firms — the same play as Harvey, the $11 billion legal AI unicorn, or Sweden's Legora, valued at $5.5 billion. The pitch was straightforward: AI drafts the first pass, lawyers finalize. Firms would pay for the efficiency.

Last September, they killed the model.

"We terminated all of our existing contracts," Ehsas told Business Insider. "Law firms and existing legal service providers are too slow and are not incentivized to use AI efficiently." The problem wasn't accuracy. It was accountability. Law firms that would use AI software remained liable for its output — which meant they had every reason to distrust a tool that could hallucinate and expose them to malpractice. "Eating glass," is how Ehsas described the experience of selling AI into a profession built on self-protection.

Moritz's answer was to skip the middleman entirely. Instead of selling software to law firms, Moritz became a law firm. It drafts contracts, employs a network of 50-plus contracted lawyers from Big Law firms including Fenwick, Cooley, and Goodwin, and carries what Ehsas calls "a very large insurance policy" on the output. When the AI makes a mistake, Moritz — not the client — is responsible (Sifted, Artificial Lawyer).

The numbers are designed to make traditional firms look broken. In its first three months of operation, Moritz says it handled contracts worth $2 billion in aggregate value across more than 100 companies in the US, Europe, and Australia. Average turnaround: four hours (Moritz). One recent engagement: a $290 million master services agreement drafted in 24 hours for a US financial services firm; the counterparty's traditional law firm took four weeks to review the same document. The contract was signed with two non-material edits (Tech Funding News).

Pricing is flat and public. An NDA costs $300. An offer letter, $500. A master services agreement, $1,500. Enterprise clients pay $500 to $700 per contract redline or iteration. No hourly billing. No surprise invoices (Moritz).

The model is what legal tech observers are calling a NewMod: an AI-native operator that owns the regulated service delivery rather than selling tools to incumbents. Ehsas has said openly that Moritz's goal is to become "the largest law firm in the world" by total contract value closed — not by headcount or revenue, but by the dollar volume of deals it helps execute (Artificial Lawyer).

Airbnb CEO Brian Chesky responded to a cold outreach from Ehsas within seven minutes, connecting him directly to the company's general counsel. The Y Combinator network, which Moritz used as its first customer base, produced trials with some of the largest companies the accelerator has ever backed (Business Insider).

Yet the model carries structural risks that the funding round doesn't resolve.

Seven engineers and operations staff are handling the volume. The contracted lawyer network is fractional — lawyers working as co-counsel on a per-matter basis, not employees. The quality control that a large law firm enforces through institutional hierarchy and partner oversight is replaced by software evaluations and a presumably large insurance policy. Moritz says it refuses to work with junior attorneys, drawing its network exclusively from lawyers with Big Law experience. Whether that gatekeeping is sufficient at scale is untested (Sifted).

There is also the liability question the entire legal AI market has tried to disclaim. Legal software companies typically include clauses warning that AI can hallucinate and that lawyers remain responsible for checking the output. Moritz's bet is that taking full liability is a feature, not a burden — that clients will pay a premium for accountability. That's a defensible position in a market where legal malpractice is rare but reputational catastrophe is not. It's also a bet that one bad contract, one missed deadline, one client who discovers their AI-drafted MSA contained a material misrepresentation, could reset everything (Business Insider).

Harvey and Legora are still selling software to the firms Moritz is trying to displace. Both have raised at substantial valuations — Harvey at $11 billion, Legora most recently at $5.5 billion with backing from Nvidia — and both have enterprise deployments inside actual Am Law 100 firms. The incumbents are not standing still. Whether the tool-selling model or the disintermediation model produces the larger business is the question the Moritz raise has made unavoidable.

Ehsas's pitch to the tech investor community landed because it translated cleanly into a founder fantasy: take your domain expertise, add AI, own the customer relationship, and skip the incumbents who would never buy from you anyway. The $9 million in four days suggests that fantasy has market price. What it doesn't tell you is whether the liability Ehsas is so eager to absorb will remain manageable when Moritz is doing a hundred $290 million deals a month instead of one.