Across firms in Mumbai, London and New York this year, the same question has been coming up in innovation committees: should we build our own AI capability, or buy it. The instinct to build is easy to understand. It gives the law firm control over its own data, no dependency on a vendor’s roadmap and the appeal of owning something outright rather than renting it. A handful of firms have actually tried, almost always by acquiring a team rather than growing one from scratch.
What’s less visible from outside is how many of them are hitting the same wall part way in the form of professional ethics rules.
In India, where I’m based, the position is unambiguous: the Bar Council of India’s rules bar advocates from entering into partnership or sharing fees with anyone who isn’t an advocate, and Indian law firms aren’t even recognised as legal entities separate from the individual lawyers who make them up. There is no ownership structure through which a non-lawyer could hold equity even if a firm wanted to offer it. It’s the same underlying idea that shows up, in different words, across most of the world’s major legal markets: a law firm cannot share equity, profit, or ownership with someone who isn’t a lawyer. In the US, this is Model Rule 5.4. It has close equivalents across most of Asia and continental Europe.
It isn’t universal, though. England and Wales opened the door in 2007, letting non-lawyer investment into what the profession calls Alternative Business Structures. Arizona went further in 2021 and abolished Rule 5.4 outright, so KPMG now operates a licensed law firm there under that framework, which is as clean a signal as exists that accounting and consulting giants see an opening the moment the rule lifts. Utah and the District of Columbia are each running their own version of the same experiment. So the picture isn’t one settled global rule. It’s a live split: most jurisdictions, India included, hold the line completely; a small but growing handful are actively testing what happens when it’s removed.
For a firm sitting on the wrong side of that line (which is still most of them) the consequence is structural, not temporary.
What the rule actually does to law firms recruiting for tech
The Code prohibits non-lawyers from holding equity, partnership shares, or profit-sharing stakes in a law firm. It exists to protect legal judgment from commercial influence, and that purpose is sound. What’s less often traced through is what it does to your position in the market for the engineers you’d need to build serious AI capability in-house.
| Compensation | Trajectory | |
| Venture-backed legaltech / tech giants: Harvey ($11B valuation), Legora ($5.55B), Microsoft, Anthropic, OpenAI, Google | Market base salary plus high-upside equity, RSUs, or stock options | Unlimited upside tied to a multi-billion-dollar enterprise valuation |
| Law firm’s internal engineering function | Cash salary plus a discretionary annual bonus, nothing more | Fixed income, no equity upside, secondary status inside the institution |
Top AI engineers, data scientists, and product architects are rarely optimising purely for base salary. They’re optimising for the equity that generates real, sometimes generational, wealth if the company they’ve joined scales or goes public. A law firm, even at $2 billion in revenue, cannot offer a single share of equity to a CTO, lead data scientist, or senior AI engineer. A startup worth a fraction of that can, and does.
The practical consequence: when a law firm sets out to build a 30-person internal AI group, it is not choosing between good engineers and slightly-less-good ones on the open market. The law firm is structurally excluded from the top tier entirely, left hiring the layer below it, or leaning on expensive external consultancies to fill the gap temporarily.
Two workarounds, examined
To pressure test options that may be suggested by innovation committees:
“We’ll route it through an ALSP subsidiary, or a UK ABS entity, where equity sharing is permitted.”
The UK’s Alternative Business Structure model, or a distinct ALSP subsidiary (Cleary’s ClearyX is a working example), does allow non-lawyer equity. But that equity sits in a boutique internal service centre, not a venture-backed company with liquid, scaling upside. The same engineer the law firm will attempt to recruit will compare the ALSP’s equity against global software company’s, and find no upside.
“We don’t need top research scientists, just integrators who can connect a few APIs.”
To be in denial about the level of expertise that moves the needle, doesn’t help. Wiring together Azure OpenAI or Anthropic’s API without genuinely senior engineering oversight tends to produce exactly what you’d expect from that description: brittle systems, hallucinations that surface at the worst moment, latency problems, and security gaps. Running enterprise-grade legal AI well requires continuous, senior engineering attention which is precisely the talent the law firm compensation structure can’t retain on a cash-only basis.
How this influences practice development
Once this constraint is genuinely understood, it changes how the build-versus-buy question should be framed internally.
Structurally, it doesn’t mean that a law firm can’t build any internal AI capability at all. A lot of the work involves integrating existing models into internal document systems, adapting a platform to the practice groups’ actual workflows, maintaining the products already licensed. This part doesn’t require the kind of frontier research talent that equity is competing for in the first place. Where the constraint bites specifically is at the top of that talent market: the engineers and researchers capable of building something genuinely novel, rather than adapting something that already exists. That tier is where equity, not salary, decides who goes where.
Which means the realistic strategy isn’t “try harder to compete for that talent” but to recognise which parts of the work only ever needed integration skill. Law firms need to build or hire for that capability with a clear conscience while treating anything that requires frontier capability as something to acquire externally. That acquisition can be through a platform, a partnership, or a structure built specifically to hold the equity the core practice legally cannot, if jurisdictional rules permit it.
This structural condition is not temporary. A pricing disadvantage narrows if you negotiate well. A feature gap closes if a competitor ships an update. This doesn’t close on its own, and it doesn’t wait for a friendlier budget cycle. It is a fact about who a law firm is legally permitted to pay, and in most of the jurisdictions this piece has surveyed, that fact isn’t currently up for debate. Planning around it now costs less than discovering it the hard way, eighteen months into a hiring effort that was never going to work.
Lawfinity Solutions advises international law firms on cross-border legal market positioning. If the India corridor is a live question for your firm, we would be interested in a conversation. Lawfinity works with one firm per jurisdiction. Engagements begin with a single conversation about your firm’s current position and where the corridor question is live for you. Write to Prachi Shrivastava
