Saudi Arabia has poured USD bns into AI infrastructure, but is the economy around it ready to use it? Counterpoint Research’s latest Global AI Cities Index ranks Riyadh 36th in the world, behind Dubai in fourth place and Abu Dhabi in eighth, a gap that can close only as adoption catches up with the hardware.
In numbers: AI adoption could boost Saudi Arabia’s real GDP growth by 0.1-0.4 percentage points annually over the next decade, the IMF estimates (pdf). Gains could reach 0.6 pp and lift GDP by up to 6% over the decade if investment scales up, AI preparedness gaps with advanced economies narrow, and AI services exports grow. However, gains depend on investments delivering expected productivity improvement.
We put five questions to Marc Einstein, research director at Counterpoint (LinkedIn), on how the market structures itself from here and where that leaves Saudi startups. Edited excerpts from our conversation:
Enterprise: How do you assess the current development of Saudi Arabia's AI market?
Marc Einstein (ME): I was at last month’s Leap conference in Riyadh, and the momentum is hard to miss. Over four days, the event saw announcements, investments and agreements worth nearly USD 15 bn across AI infrastructure, data centers, cloud computing and technology manufacturing.
Investment and infrastructure are clearly running ahead of adoption. Our 2025 Global AI Cities Index reflects this: Dubai ranked fourth in the world and Abu Dhabi eighth, while Riyadh sat at 36th, though Riyadh was among the fastest-growing AI cities in the world. When it comes to AI in government, Abu Dhabi is the gold standard in the region. It’s building an AI-native government with USD 3.5 bn committed through 2027, has already trained 95% of public sector employees in AI, and has a research base in MBZUAI and TII. Riyadh should climb sharply on infrastructure and vendor activity, but closing the gap with the UAE will depend on adoption across government and industry.
E: How does the market structure evolve from here?
ME: I expect a layered market. A few very large infrastructure players will sit at the base, with a much more fragmented application layer on top. The base is already taking shape: AWS will deploy and manage up to 150k Nvidia GPUs in a dedicated AI Zone in Riyadh, and at Leap AMD said its Instinct MI355X GPU systems are now live in Saudi Arabia alongside Cisco networking, serving Humain customers. On models, the approach is pragmatic rather than purist. Humain’s humain-m3 is an Arabic-language model based on Chinese lab MiniMax's M3, which shows the priority is sovereign control and Arabic capability, not owning every layer.
E: Does the scale of these large players squeeze local startups?
ME: Yes, particularly for startups building horizontal products such as general chatbots or foundation models, which compete directly with players that have sovereign backing and privileged access to compute. Talent is another pressure point. Funding is also more selective than the headlines suggest. According to Magnitt, Saudi Arabia led the region in 2025 with USD 1.72 bn across 257 disclosed deals, but funding fell 74% year on year in 1H 2026, while dealcount fell only 41%. Much of that drop reflects mega-rounds that didn't repeat, but the message for founders is clear: investors want traction, not just an AI story.
E: You’ve described the squeeze. What do these same investments open up?
ME: I think this is the bigger story. Large-scale infrastructure gives startups compute they don't have to finance themselves, and every enterprise buying AI capacity will need applications and integration to use it. Leap showed the support structures forming. The Digital Government Authority and STV announced a partnership under which the DGA will give participating startups access to government use cases, an experimental innovation environment, procurement opportunities, and regulatory and funding pathways. Humain is also becoming an investor, with a USD 10 bn venture fund, Humain Ventures.
E: If you were advising a founder in Riyadh today, where would you tell them to build?
ME: The strongest opportunities are in applications and enablement, not in competing on infrastructure. I'd highlight Arabic-language and culturally specific AI; vertical solutions for energy, government, healthcare, finance, logistics and the giga-projects; agentic AI that turns compute into measurable productivity; and governance, security and compliance tools as adoption scales. Fintech is a natural bridge.