🔓 The full version of this article is for premium subscribers only.
There's an old joke among corporate lawyers that the profession's core quantum leap over the past century has been the transition from yellow legal pads to white ones.
It's the kind of joke that only works because it's basically true.
Law has been, for as long as anyone working in it can remember, spectacularly resistant to technological change, and for entirely rational reasons.
Lawyers bill by the hour. Efficiency is thus, in a very literal sense, bad for business.
The incentive structure rewards thoroughness, caution, and the kind of painstaking document review that eats up bleary-eyed associate time at $500 to $800 an hour.
A single partner at a major firm can generate roughly $3.6 million in annual profit from a team of four associates billing 2,000 hours each.
Show me the incentive, as the saying goes, and I'll show you the outcome.
A startup selling a time-saving tool into that business model is, whether it knows it or not, asking the firm to give up revenue.
And then LLMs arrived, and something genuinely strange happened – the formerly most conservative profession in the knowledge economy turned out to be the single best commercial use case for generative AI.
It should never have been a surprise really, as Sequoia's Bogomil Balkansky characterises it, the profession is built on "text in, text out."
Contracts, briefs, memoranda, regulatory filings, due diligence reports, board resolutions, settlement agreements, all of it is written language with structure and precedent, which is exactly what these models were built to process.
As for the pricing paradox, Harvey's CEO Winston Weinberg recently told the 20VC podcast that firms are now winning mandates by building custom Harvey workflows into their pitch for M&A deals. The argument, essentially, is that AI expands the capacity for legal work rather than cannibalising it, negating the billable hour conundrum. For in-house teams, the calculus is simpler still: every hour saved is a direct cost reduction.
What has followed is a land grab moving at a speed that the legal industry has never experienced and is not entirely equipped to process.
Harvey AI, founded in 2022 by a former litigator and a former DeepMind researcher who were flatmates in LA, has raised nearly a billion dollars across seven rounds and reached an $8 billion valuation in December 2025.
It crossed $100 million in ARR in August and hit an estimated $190 million by year-end.
🔓 The full version of this article is for premium subscribers only.
The champion across the Atlantic is Legora, a Stockholm startup founded by a 25-year-old Swede, which has gone from YC to reportedly eyeing a $5 billion-plus valuation in roughly two years, raising Europe's largest-ever legal tech venture round in the process.
These are the two big dogs you're already most likely familiar with from funding headlines and extensive podcast circuits but they are by no means alone – overall, global legal-tech funding had one hell of a year, reaching $3.2 billion in 2025.
And then, in early February, the foundation model companies themselves started arriving, with Anthropic launching a legal plugin for its Claude Cowork platform that automates contract review, NDA triage, and compliance tracking, configurable to an organisation's own playbook.
The impact on public market confidence in legal tech incumbents was seismic.
RELX, the parent company of LexisNexis, posted its steepest single-day share price fall since 1988.
Three weeks later, Anthropic did the same thing to cybersecurity with Claude Code Security, and CrowdStrike lost 18% of its market cap in days.
The companies that build the models are moving into the application layer, and the startups wrapping their products around those same models are being squeezed from both sides.
And with all that being said, Harvey's own CEO, Winston Weinberg, said during a Reddit AMA in December that there are around 10 million legal professionals globally and Harvey serves just single-digit percentage points of them.
The global legal market is worth an estimated $1 trillion, he said, but only about $30 billion of that is currently spent on tech.
Legal tech penetration in other words sits at roughly 3% if we take his estimate at face value.
The opportunity, we can infer, is so large that even the most aggressively funded company in the space describes itself as barely scratching the surface.
Why is it then that in a year where MENA venture capital hit $3.8 billion in 2025 according to MAGNiTT, with AI alone capturing $858 million, legal tech has captured effectively nothing in the region?
Shouldn't investors be falling over themselves to unearth the next serial founder team who can adapt a model clearly working exceptionally in the West to the local regional context?
Is legal tech in the AI era, to coin a phrase, uncopycatable?
What is it exactly about the legal arena, unlike e-commerce, fintech, food delivery, and proptech that is failing to capture the imagination of the GCC capital deployment class?
Why law is different here
The short answer is that law, unlike nearly every other sector the MENA startup ecosystem has successfully adapted from the West, has an architectural problem that even the Samwer brothers would struggle to overcome.
You don't get ahead by reading what everyone else reads.
Weekly deep dives, founder and VC playbooks, the Careem Mafia Database, the MENA Tech IPO Tracker, the Diaspora 50, and two years of archived analysis
