General Legal Is Rebuilding the Law Firm to Scale Like Software
Ask any startup founder about their lawyers and watch the expression change.
It's not that the work is bad. It's that the model is maddening. A contract lands in your inbox from a customer or a vendor, and you need it reviewed before you can sign. So you email outside counsel. You wait. The clock is running at $500 to $2,000 an hour, billed in six-minute increments, and you have no idea what the final number will be. Days pass—sometimes weeks. The deal loses momentum. And when the redline finally comes back, you're never quite sure whether the lawyer read every clause closely or skimmed a familiar template and moved on to a bigger client.
For a fast-growing company, this is a genuine bottleneck. Every unsigned contract is stalled revenue, a delayed hire, or a partnership on hold. The legal function—which should be an enabler—becomes a chokepoint measured in billable hours.
This is the problem General Legal is built to solve.
General Legal is an AI-native law firm for growth-stage companies. It pairs proprietary AI workflows with real, US-barred attorneys to review, negotiate, and draft commercial contracts—at a flat fee, delivered in hours instead of weeks. Their tagline captures the ambition precisely: outside counsel that scales like software.
Why Legal Work Resists Automation—and Why That's Changing
Legal has been one of the most stubbornly un-automated professional services, and for understandable reasons.
The work is high-stakes and adversarial. A missed indemnification clause or an overlooked liability cap can cost a company far more than the contract was ever worth. The language is dense and context-dependent; the same phrase can be benign in one agreement and dangerous in another. And critically, the work carries professional and legal accountability. When an attorney signs off on a contract, they're staking their bar license and their reputation on it. That accountability is not a formality—it's the entire product. It's the reason a client can act on the advice with confidence.
This is exactly why "just use an AI" has been a dead end for serious legal work. Large language models are extraordinary at reading a contract, summarizing it, spotting patterns, and flagging the clauses that usually cause trouble. They're also capable of confidently asserting things that are wrong—the failure mode that is completely unacceptable when the output is legal advice a company will rely on. Nobody wants to sign a seven-figure agreement on the say-so of a model that might be hallucinating.
General Legal's insight is that this isn't a reason to keep AI out of legal work. It's a reason to design the workflow correctly.
The AI-Plus-Attorney Workflow
General Legal's model puts AI and human attorneys in the roles each is actually good at.
The process is deliberately simple from the client's side:
Submit. A client sends a contract through a dedicated Slack channel, email, or the client portal—wherever they already work.
Quote. They get a flat-fee price upfront. Not an hourly estimate that balloons, but a fixed number they can approve before any work begins.
AI triages. Proprietary AI agents read the document, summarize it, triage the issues, and flag the risky terms—doing in minutes the first-pass analysis that would otherwise consume an associate's afternoon.
An attorney delivers. A US-barred attorney reviews the AI's work, applies judgment, and delivers the final product under their own professional license. As the firm puts it bluntly: no AI slop. A human lawyer stands behind every deliverable with their name and bar number attached.
The economic magic is in the division of labor. AI eliminates the expensive, repetitive first pass—the reading, summarizing, and flagging that traditionally eats the bulk of billable hours. That collapses both the cost and the turnaround time. But the attorney remains the accountable party, which preserves the one thing clients are actually paying for: the confidence to sign. The AI makes the work fast and cheap; the lawyer makes it trustworthy.
The results show up in the numbers General Legal advertises. A standard contract review runs a flat $500 rather than a variable four- or five-figure bill. Simple reviews start lower; full negotiation and drafting are priced as fixed tiers. Median first turnaround is under three hours. And the firm reports representing hundreds of growth-stage companies with a customer satisfaction score in the mid-90s—the kind of retention signal that's hard to fake in a business where the work either holds up or it doesn't.
Flat Fees Change Everything
It's easy to skim past "flat fee" as a pricing detail. It's actually the strategic core of the whole model.
The billable hour is the defining institution of the legal industry, and it quietly misaligns lawyer and client at every turn. Under hourly billing, inefficiency is revenue. The faster a firm works, the less it earns. There's no structural reward for adopting technology that would let an attorney finish in one hour what used to take five—if anything, the incentive runs the other way. This is a large part of why Biglaw has been so slow to automate: the business model punishes the very efficiency that AI enables.
A flat fee inverts the incentive. When General Legal charges $500 for a contract review regardless of how long it takes, every minute the firm saves through AI flows to its own margin. Efficiency becomes the profit engine instead of the enemy of it. The firm is now motivated to build the best possible AI tooling, because better tooling directly improves the economics of every job.
For the client, the flat fee removes the two things they hate most: unpredictability and the perverse sense that they're being penalized for a lawyer's slow morning. They know the price before they commit, and they know it won't move.
This is why an incumbent firm can't simply copy the approach. Adopting flat-fee, AI-accelerated work would cannibalize the billable hours that are the foundation of their revenue and their partner compensation. General Legal has no such conflict. It was built from a clean sheet around the new economics.
Founders Who Have Lived Both Sides
The team is a large part of why this is credible, because you cannot build this company from only one discipline.
You need deep legal expertise to know which clauses matter, how negotiations actually unfold, and where the professional-responsibility lines are drawn. And you need to have built AI-native legal software before, because the hard part is engineering a workflow where the AI accelerates without ever compromising the attorney's accountability.
General Legal has both. Its founding team pairs Harvard Law attorneys who practiced at top firms like Fenwick and Cooley—people who lived inside the Biglaw model they're now trying to unbundle—with a technical founder who was CTO of Casetext, one of the defining legal-AI companies of the last generation, through its acquisition by Thomson Reuters. That's a rare combination: the people who know exactly what elite commercial legal work requires, working alongside someone who has already built AI that lawyers trust and shipped it at scale.
This matters because the failure modes here are subtle. Build it too "AI-first" and you produce fast, cheap output that no serious company will rely on. Build it too "lawyer-first" and you've just recreated a traditional firm with a chatbot bolted on. Threading that needle requires founders who genuinely understand both worlds.
The Larger Unbundling
Zoom out, and General Legal is an early instance of a pattern that's going to repeat across every professional-services industry.
Law, accounting, consulting, and similar fields share a structure: expensive experts, billing by the hour, doing a mix of genuinely high-judgment work and a great deal of routine analysis that merely requires expertise to supervise. AI is exceptional at that routine analysis. It's the reading, the summarizing, the first-pass flagging, the drafting from a known template. What it can't do—yet, and maybe ever, in the way clients need—is bear accountability. It can't put a license on the line.
The winning model in every one of these fields is likely to look like General Legal's: AI does the analytical heavy lifting, a credentialed human provides judgment and accountability, and the pricing shifts from hourly to fixed because efficiency is now something to be captured rather than something to be avoided. The firms that thrive won't be the ones that ban AI to protect their billable hours, nor the ones that replace professionals with a raw model. They'll be the ones that redesign the workflow around what AI and humans are each actually good at.
General Legal is running that playbook in commercial contracts—one of the highest-volume, most standardizable, most universally-hated corners of legal work. It's a smart place to start: the pain is acute, the work is repeatable, and the customers are exactly the fast-moving growth companies who feel the old model's friction most sharply.
What a Law Firm Becomes
For a century, the law firm has been a fundamentally human institution whose capacity scaled linearly with headcount. To do more work, you hired more lawyers. To grow, you added associates and partners. The economics, the culture, and the billable hour all followed from that basic constraint.
General Legal is trying to break the link between capacity and headcount. If AI handles the first pass on every contract, an attorney can supervise and deliver far more work than they ever could alone—without lowering the quality bar, because their judgment and license still sit behind every result. The firm's capacity starts to scale like software: with tooling, not just with hiring.
That's what "outside counsel that scales like software" really means. Not a chatbot that dispenses legal tips, and not a cheaper version of the same slow model—but a genuine reimagining of how legal work gets produced, priced, and stood behind.
If it works, the winners won't just be a faster, cheaper alternative to the traditional firm. They'll have changed what a law firm is.