The cap table is the country

Lebanese founders are raising real money in artificial intelligence. The gap between what they raise abroad and what they raise at home is now a number you can check.

Rayan Dabbagh left Beirut at seventeen. Grand Lycée Franco Libanais, then Georgia Tech, then Amazon Web Services, then out on his own. This year he launched Bounty, a marketplace in New York where artificial intelligence agents claim paid work, complete it, and get released the money only once a result has been verified. Andreessen Horowitz put in a million dollars.

The company was months old.

In February of this year, HAQQ Legal AI raised three million dollars in Beirut. It is a serious business: a legal engine that produces output under the law that actually governs the question, practice management and payments built on top, more than seven thousand clients including bar associations and courts. Six years of work. Wamda and L’Orient Today both reported the round.

Three million dollars is the largest confirmed artificial intelligence round in the country.

Hold those two figures next to each other, because the comparison is the whole argument. The best that six years and a real client base can raise inside Lebanon is three times what a New York fund will write on a first cheque to a company that is a few months old. Not because the Beirut company is worse. Because of where the cap table sits.

It has become a national habit to talk about brain drain as though the loss were the people. It is not. The people are doing extremely well.

Karim Atiyeh was born and raised in Lebanon. He sold his first company to Capital One in 2016, co-founded Ramp in 2019, and is now its Co-CEO. Most Lebanese coverage still calls him the chief technology officer, which has been out of date for a while, and the fact that nobody here noticed the change is itself part of the story.

Nadine Nehme took a doctorate under a Nobel laureate and co-founded Medicus AI, which now processes medical data for more than thirty five million patients across eighteen markets. The company is routinely described as Lebanese. It was founded in Dubai in 2015 and moved its headquarters to Vienna the following year. Beirut is one of five offices.

Elie Habib co-founded Anghami, took it onto NASDAQ, and then this year built something called World Monitor in a single day because he wanted it for himself. It maps world events in real time from more than four hundred and fifty sources, with a language model doing the final classification. Within weeks it had four million visitors from a hundred and seventy four countries. It is open source. He gave it away.

That is not a country losing its talent. That is a country whose talent is compounding somewhere else, on somebody else’s balance sheet.

What leaves is not the person. It is the equity.

When Dabbagh raised in New York, the ownership went to a fund on Sand Hill Road. When Ramp is eventually valued, the gain sits with investors who have never been to Beirut. When Medicus was priced, it was priced in Vienna. The work is Lebanese. The upside is not, and the upside is the part that builds anything permanent.

This is the same arithmetic that governs remittances. Five point eight billion dollars reached Lebanon in 2024, nearly eighteen per cent of the economy, and almost none of it arrived as investment. It arrived as maintenance. It paid school fees and medical bills and kept households standing, which is honourable and urgent and does not compound. A remittance is spent once. Equity keeps paying.

So the diaspora sends money home and keeps its ownership abroad, and both halves of that sentence are rational for the individual and ruinous in aggregate.

There is now a fund. B&Y Venture Partners has launched a five million dollar vehicle aimed at Lebanese startups in artificial intelligence, financial technology and software.

Five million dollars is not a large fund. Andreessen Horowitz has written single cheques larger than the whole of it. But it is the first time in years that a founder in Beirut has had a domestic address to send a deck to, and the direction of that money matters more than the size of it.

What is missing is not capital, in the end. It is the register. A fund in Beirut with five million dollars and a founder in New York with a million from a top tier firm are, in principle, useful to each other. In practice neither can reliably find the other, because there is no maintained record of who is doing what and where.

That absence is expensive in a way nobody bills for. It is why the same network gets rebuilt from zero every twenty five years, and why a Lebanese engineer arriving in Dubai next month cannot reach the one who arrived fifteen years ago.

Our companion, Club Cedro, has begun keeping the record for artificial intelligence specifically: who is building, where they are based, what the money actually is, and where the evidence for each claim comes from.

It found that most of what circulates about Lebanese artificial intelligence is not checkable. The lists everybody quotes trace back to a handful of directories and, in one case, to a coding school’s marketing blog whose funding figures are wrong by a factor of nine. Companies that are genuinely here are missing from them. Companies that left years ago are still listed as Lebanese.

Nine entries survived verification. Twenty one were opened and left off, and the reasons were published rather than kept quiet.

That is a small number and it is meant to be. A record you can check is worth more than a list you cannot, and the only version of this that has ever been worth anything is the version where somebody is willing to publish what they excluded.

THE ASK, WHICH HAS NOT CHANGED

Lubnan will never ask you for money. There is no fund here, no fee and no vehicle to pay into.

The ask is the same one it has always been, and this piece is only evidence for why it matters. Be findable. Let your name sit on a record that says what you built and where, so the fund in Beirut can find the founder in New York without either of them relying on an accident. Take one call a year from somebody thirty years behind you. Name the people you know who belong on the record and are not on it, which is still most of them.

The talent question was settled a long time ago. What is unsettled is whether any of it ever accumulates in one place long enough to become an institution.

That is not a question about money. It is a question about whether we bothered to write anything down.