I want to start with optimism, because it is earned. Spend any time around the Kingdom's research institutions, its funds and its founders, and the energy is unmistakable. Something is being built here, deliberately and at speed, and the people building it are serious. The pace of change is real, and anyone who waves it away has not been paying attention.
So here is my honest read on where we are. I want to focus on one thing in particular: venture building. By that I mean the unglamorous, difficult business of turning research and ambition into actual companies. Because that, more than any strategy document or funding announcement, is where the next chapter of the Kingdom's deep tech story will be written.
The framework is built
Let me give credit where it is plainly due. The top-down work has been done, and much of it has been done well. Vision 2030 set the direction, the pace and the targets, and a generation of agencies, funds and regulatory reforms has been stood up behind it. The scaffolding for a deep tech ecosystem now exists, and in several respects it is as good as anything I have seen anywhere. If you had shown me the 2026 landscape back in 2022, I would not have believed how far it would travel in so short a time.
That part, the framework, is not the problem.
A framework is not an engine
Here is the honest bit. A framework is not the same thing as an engine, and the two are easy to confuse. The scaffolding is excellent. The dynamic mass that is supposed to fill it, the founders, the university spinouts, the dormant IP finally being pulled out of the lab and commercialised, is not yet at the level the framework deserves.
A framework is not an engine.
I say this from observation rather than from a spreadsheet, but the pattern is familiar. Compared with the two markets I know best, London and Austin, the Saudi scene is still heavy with enablers, ministries and regulators, and lighter on the sheer founder density that makes an ecosystem hum. Some of that is simply lifecycle: these things take time, and the Kingdom started later than most. Some of it is cultural, the long pull of the state and the old industries, and a relationship with risk and failure that does not change overnight. None of this is a criticism. It is simply where the work now is.
The question has changed
The most important shift is not in the ecosystem itself. It is in the question we should be asking about it.
In 2022 and 2023, the right question was: is the framework there? Today, in 2026, that question has been answered, and a harder one has taken its place: is the framework producing companies? The bottleneck has moved from strategy to execution, from building the ecosystem to building ventures inside it.
The pace of change, in numbers
| Indicator | Then (2022 / 2023) | Now (2025 / 2026) |
|---|---|---|
| Global AI Index | 31st in the world overall (2023) | 14th overall, and No. 1 in the world for AI government strategy (2024) |
| Startup ecosystem rank | Climbing, but mid-table globally | Up 27 places to 38th, the biggest jump of any country; StartupBlink Country of the Year 2025 |
| Riyadh, as a city | Outside the global top 100 for growth | Fastest-growing city in the global top 100, at +134% in a single year |
| Venture capital | Leading MENA into a cooling market | $750m deployed in 2024, then $860m in H1 2025 alone, up 116% year on year |
And deep tech ventures are genuinely hard to build. They carry long R&D cycles, heavy intellectual property, real regulatory weight and a hunger for capital, and they do not assemble themselves out of a research paper and a good intention. They need a structured path from the lab to a company. Just as importantly, they need to be selected and backed on evidence, on what a venture can actually demonstrate, rather than on the quality of a pitch or the warmth of an introduction.
Why this matters, and who has to do the lifting
This matters because deep tech is the competitive advantage that Vision 2030's diversification actually depends on. Diversification without competitive advantage is just a wider set of things to be average at. Deep tech, being R&D-rich, IP-rich and hard to copy, is how a knowledge economy earns its edge.
The Kingdom has one real advantage here: a relatively clean slate. You can design the venture-building layer properly from the start rather than retrofitting it later. But, and I have said versions of this for years now, the heavy lifting has to shift from the state to commercial actors. The state has played the architect superbly. The next phase belongs to venture builders, investors and operators who can run agile, evidence-led processes down at the level of the individual founder and the individual company. The architect frames the building. It should not also have to be the engine inside it.
What good looks like
If I had to compress my view into a handful of practical points, it would be these. Convert research into ventures systematically, with a real researcher-to-entrepreneur process, rather than hoping that a brilliant scientist will spontaneously become a founder. Select with rigour and evidence rather than instinct, judging teams on what they can demonstrate, not on how well they pitch. Back ventures on that same evidence, and put real resources behind the hardest moment in the journey, the leap from a working prototype to a deployable product, where most deep tech dies. And let commercial actors run it: agile, founder-level, evidence-led, with the state continuing as the architect rather than trying to be the operator as well.
None of this is a criticism of what has been built. It is the opposite. The framework is good enough that the next conversation can finally be about the engine. That is a sign of progress, not a complaint about it.
This is the gap The Innovation Races is built to close: a structured path from the lab to a company, and selection made on evidence rather than instinct. Explore The Innovation Races, or get in touch.