Law is the gateway drug

Here's a number that should stop people in their tracks.

Harvey, the most prominent legal AI company, was valued at $11 billion at its last funding round. That's roughly ten times the largest accounting AI company. No standalone consulting AI company exists at meaningful scale. In a landscape where everyone agrees AI will transform professional services, the serious money — the money that represents a bet on the future — is overwhelmingly flowing into law.

Why?

The easy answer is that law is text-heavy, precedent-based, and therefore a natural fit for language models. That's true but insufficient. Accounting is also text-heavy and rule-based. Consulting is also knowledge work. The "AI is good at legal text" explanation doesn't account for a 10:1 valuation differential. Something deeper is going on.

We think the answer lies in law's structural position — a position that the profession has built, quite strategically, over the course of a century. And understanding that position is the key to understanding everything else that follows in this series.

The Hundred-Year Strategy

David has spent decades studying the legal profession as a system. And the pattern, once you see it, is striking: lawyers have worked for a hundred years to make themselves a central node in how modern society operates.

This wasn't any individual's strategy. No one sat in a room and planned it. It was a collective, emergent one — the legal profession steadily, often brilliantly, positioning law as the centerpiece around which business, public policy, and everything else has to operate.

Think about what it means to be a General Counsel today. You're not just answering a legal question. You're increasingly at the heart of every major business decision. You're at the closing table. You're in the board meeting. You're on the call when the CEO is deciding whether to proceed with an acquisition, enter a new market, restructure the company, or respond to a regulatory investigation. The lawyer's opinion isn't just one input among many — it's often the gating input, the one without which nothing moves forward.

How did law get here? Several mechanisms, operating over decades.

Start with regulation. Lawyers didn't just respond to it — they helped create it, and they shaped it in ways that require legal interpretation. Securities law, antitrust, employment law, tax, corporate governance — each domain generates complexity that creates demand for the very profession that helped write the rules. This isn't conspiracy. It's professional ecosystem design.

Then there's transactional centrality. Corporate transactions — mergers, acquisitions, IPOs, financings — are structured as legal events. Not because they have to be (a merger is fundamentally a business combination), but because law has positioned itself as the framework through which these events achieve legitimacy and enforceability. The deal doesn't close without the opinion letters. The money doesn't move without the legal documentation. The signatures don't happen without the lawyers at the table — or increasingly orchestrating DocuSign.

At the turn of the twentieth century, the modern "Cravath-style" law firm was the first to solve a problem that other professions haven't. In professional services, reputation is everything — but reputation had traditionally been attached to individuals, which makes it portable and fragile. What Cravath realized was that a Cravath (or Sullivan & Cromwell or Davis Polk) opinion letter didn't just carry the individual lawyer's judgment. It carried the firm's institutional weight. That opinion letter functions as a currency of trust in the capital markets, and it's very hard to replicate outside the profession. This creates a structural lock-in that accounting and consulting haven't achieved to the same degree — which helps to explain why top accounting and consulting firms like Arthur Andersen and McKinsey initially modeled themselves to look like Cravath-style law firms.

And lawyers serve as gatekeepers not just for their clients but for the system. Regulators rely on law firms to certify compliance. Courts depend on lawyers to frame disputes properly. Boards delegate to counsel the certification that proper processes were followed. This gatekeeper role gives lawyers access to — and influence over — virtually every major institutional decision.

The result is a profession that doesn't just serve business. It is woven into the fabric of how business operates. An accountant can tell you whether the numbers work. A consultant can tell you whether the strategy is sound. But the lawyer sits at the intersection of both, plus regulation, plus governance, plus risk — and the lawyer's sign-off is often the final prerequisite for action, both inside the organization and in the world.

The best lawyers have understood this. What they're really providing is not technical legal knowledge. It's judgment around a complex intersection of law and business and strategy and politics and public relations. That intersection is where the value lives. And law has spent a century making sure you can't get to the intersection without passing through it.

The Legalization of Everything

If law's structural centrality were static — the same domains, the same regulatory landscape, the same business complexity — it would still matter enormously. But something else is happening simultaneously: the scope of what gets "legalized" is exploding.

Sustainability. Cybersecurity. Geopolitical risk. New technologies. Inequality. Migration. Governance. AI itself. Each of these started as a policy question, a business question, or a social question. Each is increasingly being channeled through legal frameworks. And yet law is only part of addressing these critical issues — which makes the lawyer's position even more central, not less, because someone has to integrate the legal dimension with everything else.

Consider what's landed on in-house lawyers' desks in the last five years. The EU's AI Act. The SEC's climate disclosure rules. GDPR enforcement actions running into the hundreds of millions. Sanctions compliance as geopolitics fragments into rival blocs. Supply chain due diligence legislation. ESG reporting mandates in Europe alongside anti-ESG backlash legislation in American states. Each generates new regulation, almost none of it clear or consistent across jurisdictions.

These issues are falling onto the desks of in-house lawyers not because they're purely legal problems — they're not — but because they carry legal consequences. And in a volatile, uncertain, complex, and ambiguous world, legal consequences generate legal work. David has described AI itself as being "a welfare act for lawyers" — at least in the short term, because everything from disputes over intellectual property, liability for hallucinations, and the reach and interpretation of new AI regulations will be litigated on all sides.

The legalization trend is self-reinforcing. As more domains are channeled through legal frameworks, more organizations need lawyers to navigate them. As more lawyers engage with these domains, they develop the expertise that further embeds law's relevance. The regulatory machinery generates demand for the profession that operates it. It's a flywheel. And it's accelerating.

Lawyers are being asked to integrate their knowledge of law and legal strategies into wider "integrated solutions" that extend far beyond law — business strategy, compliance, product development, government relations. Law is the entry point, but the work doesn't stay purely legal. It radiates outward into the full complexity of the client's problem. This means lawyers today need to be multidisciplinary problem solvers, not just technically proficient legal analysts. Thinking like a lawyer has to be thinking like a problem solver — thinking about being an empathetic listener, about working collaboratively with people who have different orientations and perspectives. People who traditional lawyers still foolishly call "non-lawyers," but who are now essential to every aspect of legal work.

Why This Makes Law the Gateway Drug

Now connect these two forces — structural centrality and accelerating legalization — and a picture emerges that explains the 10:1 valuation differential and much else besides.

AI that operates effectively in legal contexts doesn't just gain access to legal work. It gains access to everything law touches. And law touches everything.

When an AI platform builds a matter-centric workspace for a merger, it's not handling a purely legal task. It's navigating the regulatory analysis, the compliance review, the governance assessment, the risk evaluation — or at least it potentially can be if done properly. When a law firm and an investment bank and a consulting firm work side by side in an AI-powered shared workspace on an M&A deal, the AI isn't operating in a legal silo. It's operating at the intersection of law, business, and strategy that law has spent a century building.

That's why law is the gateway drug. An AI system that earns trust in legal contexts — where accuracy matters most, where errors carry consequences, where governance and security are non-negotiable — has demonstrated the capabilities needed to operate across professional services more broadly. And because lawyers sit at the intersection of so many domains, the legal use case naturally, almost inevitably, expands into adjacent territory.

Gateway drugs are powerful precisely because they don't stay in their lane. They open a door that leads to other doors. Law is the same. Once AI enters through the legal gateway — once it's embedded in the workflows, trusted by the practitioners, integrated into the systems — it doesn't stop at legal work. It follows the same paths that lawyers themselves follow: outward from the legal question into business strategy, compliance, risk, governance, and deal execution. The full complexity of the client's world.

This isn't just theory. We've already seen contract analytics morph into contract lifecycle management, moving from the legal department's reviewing contracts to sales and procurement where contracts are made, to risk and strategy where the data that lives in contracts is mined to reduce risk and find new opportunities. And the market is already pricing it. Investors aren't valuing Harvey as a legal AI company. They're valuing it as a potential entry point into professional services AI — because that's what law's structural position makes possible. The 10:1 ratio over accounting AI isn't about the relative size of the legal market. It's about the relative centrality of law's position and the adjacencies it unlocks.

The Three Paths

More than twenty years ago, David was teaching a class on technology and law for the very first time. A researcher came in who had studied the impact of computer-aided design and manufacturing on architecture and engineering. The first thing he said was: "Professionals always think that technology allows them to do what they already do cheaper and faster." Instead, what technology ultimately does is redefine what it means to be a professional.

Lawyers are still mostly on the first path — using AI to do what they already do cheaper and faster. Document review, legal research, first-pass drafting. This is valuable but limited. If you're doing the same thing more efficiently, why should anybody pay you more for it?

The second path involves what economists call the Jevons paradox — making legal work dramatically cheaper could increase total demand for it. Problems that were previously too expensive to address become addressable. A small business that couldn't afford to have its contracts reviewed might now be able to. A nonprofit that couldn't navigate regulatory complexity might now have access to AI-powered legal guidance. This expands the pie but doesn't fundamentally change what lawyers do.

The third path is the one that matters — and the one that law's gateway position makes uniquely possible. These tools could allow lawyers to do fundamentally different things. Not just answer legal questions faster, but serve as the central node in creating integrated, multi-disciplinary solutions to complex problems that are increasingly legalized but where law is only one piece of what the client is looking for.

This is the path where "law is the gateway drug" carries the furthest. If lawyers can use AI not just to accelerate their existing legal work but to expand into the full scope of problems their structural position gives them access to, then AI doesn't just improve law — it transforms the profession into something broader, more central, and more valuable than it has ever been.

But most lawyers are nowhere near even understanding this challenge, let alone working it through. The profession trains people to think within legal categories, not across them. It rewards deep specialization, not integrative judgment. The problems the profession is now grappling with — problems at the intersection of law and technology and geopolitics and governance — are so novel that there are no experts. It used to be that you had to work your way up from being the 10th person on the team, to being the 5th chair, then the 3rd, before one day becoming the leader. Now everybody's dealing with problems that are novel, at least in their presentation. Expertise can and should come from everywhere.

The Uncomfortable Implications

The same structural position that makes law the natural entry point for AI means law is where disruption concentrates first. Every alternative provider, every AI-native firm, every ambitious technology company targets law first — because that's where the gateway is. KPMG sought and received permission to practice law in Arizona. DIY legal startups offer wills, real estate, and dispute resolution directly to consumers. Y Combinator's 2025 Request for Startups challenged founders to "start your own law firm, staff it with AI agents." The gates are under siege precisely because they guard such valuable territory.

There's also the question of who controls what passes through the gate. Law built its centrality through regulation, transactional embedding, and the gatekeeper function. But what happens when AI platforms — not law firms — become the infrastructure through which legal and adjacent services are delivered? The centrality remains, but the question of who captures the value shifts. Platform economics have a way of concentrating value in infrastructure providers, not the professionals who use them. Law's gateway position attracted the AI investment. But the infrastructure being built could end up routing around the profession rather than through it.

Then there is the danger of AI killing the goose that lays the golden egg: if AI enters through law and automates the junior work through which lawyers have historically been trained, the gateway drug doesn't just open doors — it erodes the foundation that made the gateway valuable in the first place. The commodity legal work that AI targets first — document review, legal research, due diligence — is also the apprenticeship through which domain expertise is built. Without that expertise, who will exercise the judgment that sits at the intersection? Who will know when a plausible-sounding AI output is subtly, consequentially wrong?

The profession may be consuming the very capability pipeline that makes its gateway position viable. We explore this in depth later in the series, but the dynamic is already embedded in the gateway position itself.

And even if all of this can be navigated, there still is the fact that a "gateway drug" is still a drug. For many, the increasing centrality of law is something to be mourned not celebrated. If AI allows law to become even more important, and lawyers ever more central, do we risk creating Grant Gilmore's vision of hell in which "there is nothing but law, and due process will be meticulously observed"? Worse yet, will we produce a world in which law is nothing more than the "Code" that Larry Lessig warned us about almost thirty years ago?

Dan Wang's 2025 book Breakneck offers a comparative lens on this shadow side. Wang contrasts the United States as a "lawyerly society" — where lawyers from Harvard and Yale go not just to legal positions but to many positions in government and elected office — with China as an "engineering state," where engineers occupy equivalent positions of power. He argues this has consequences: China has been able to build bridges, roads, and manufacturing capacity at extraordinary speed, while in America, lawyers often stop things from being built.

Being at the center of everything means you can also be an impediment to everything. The gateway drug doesn't just enable — it can also obstruct. That duality — enabler and impediment, strategic asset and cost center, gateway and gatekeeper — is one the profession has never fully resolved.

What Comes Next

Where does this leave us?

Law is the gateway drug for AI into professional services because of a structural position that lawyers have built over a century. That position — being woven into the fabric of how business, regulation, and society operate — means AI that enters through law gains access to everything law touches. And the legalization of everything means that what law touches keeps expanding.

This creates the largest AI investment opportunity in professional services, and the market is pricing it accordingly. It also creates the possibility of the third path — lawyers using AI to expand from legal advisors into multi-disciplinary problem solvers at the center of the most complex decisions organizations face.

But the gateway metaphor cuts both ways — as does the drug. Law is where the transformation starts. Law is also where the disruption concentrates. And the profession's structural position — the thing that made law the gateway in the first place — could itself be transformed by what passes through the gate. What emerges on the other side could be "a nation of code, not men."

In our next piece, Harvey's Strategic Evolution, we look at the company that appears to have understood law's gateway position most clearly: Harvey AI. Its five-phase evolution — from broad legal assistant to enterprise trust layer to platform to infrastructure for multi-disciplinary collaboration to ecosystem host — reads like a systematic strategy to become the distribution system for AI across professional services, with law firms as the central node. If law is the gateway drug, Harvey may be building the pipeline.