Beyond Data Center Hosting: What AI Infrastructure Really Means for the GCC Knowledge Economy

As cloud regions, AI campuses, and data centers spread across the Gulf, the deeper question is no longer only where compute sits, but whether the GCC can turn physical digital infrastructure into GDP gains, local capability, domestic firms, and lasting knowledge-economy depth rather than simply facilitating the regional reach of global platforms.

The Gulf’s data center expansion is often framed as a story of cloud infrastructure, artificial intelligence readiness, and national digital ambition, but that account remains incomplete unless it asks a second and more difficult question: what kind of economy is actually being built around this new infrastructure? A data center can arrive in the GCC long before a true knowledge economy does. Server halls, cloud regions, and AI campuses may improve latency, attract global firms, and support digital transformation, yet they do not automatically create deep local capability, nationally rooted talent, domestic innovation, or sustained value capture. In the Gulf, this distinction matters especially because governments are not simply trying to host global infrastructure; they are trying to diversify their economies, strengthen national labor participation, develop technical capability, and build more durable forms of long-term competitiveness.

This Part II places the data center boom in that fuller economic context. It argues that the real issue is not only whether the GCC can attract digital infrastructure, but whether it can convert that physical presence into local firms, higher-value work, ecosystem depth, and genuine knowledge-economy formation. That is why we recommend moving beyond GDP headlines, job counts, and promotional narratives into the harder terrain of value capture, capability formation, benchmarking, nationalization, diffusion, and economic embeddedness. Part II is the second of a two-part series examining the deeper implications of the GCC’s data center boom. Part I focused on the physical and institutional realities of electricity demand, cooling, water stress, infrastructure logic, and governance. Part II turns to the economic and human-capital question of whether this buildout produces a genuine knowledge economy or simply hosts global digital infrastructure while leaving too much of the highest-value activity, talent formation, and innovation capacity elsewhere.

Reading Time: 40 min.

All illustrations are copyrighted and may not be used, reproduced, or distributed without prior written permission.

Summary: This Part II examines whether the GCC’s expanding base of data centers, cloud regions, and AI-related infrastructure can generate a true knowledge economy rather than merely a hosting economy. It argues that digital infrastructure is a necessary but insufficient condition for deeper economic transformation. Saudi Arabia, the UAE, Qatar, Bahrain, Kuwait, and other Gulf states may capture real GDP gains, direct investment, and service improvements from data center growth, but those benefits do not automatically translate into local capability, national talent development, domestic firm formation, or long-term strategic ownership of higher-value digital layers. We therefore distinguish between hosting value and embedded value, and introduce a practical benchmarking framework to evaluate local value capture, talent localization, capability deepening, domestic firm participation, innovation spillovers, and national embeddedness. The analysis also refines these benchmarks for the GCC context by distinguishing between local labor participation and national labor participation, recognizing that resident expatriate talent and nationally rooted capability are not the same thing. Its central conclusion is that the Gulf has a serious opportunity to turn digital infrastructure into a distinctive and durable knowledge-economy model, but success will depend on whether data centers become connected to local firms, local talent ladders, applied innovation, state strategy, and wider economic diffusion rather than remaining impressive but relatively shallow islands of imported capability.

The overall arc is: visible data center expansion and digital infrastructure growth → the deeper question of who captures value, develops skills, and owns strategic capability → GDP contribution and job creation as only the surface layer, with embedded capability, talent formation, and domestic firm participation as the more revealing measures underneath → localization versus nationalization as a distinctively Gulf issue that complicates easy claims about local impact and demands closer attention to citizen participation, retention, and upward mobility → foreign platforms, commercial scale, and globally mobile expertise as drivers that can speed infrastructure buildout while also constraining how much value becomes nationally rooted → legitimacy as something different from hosting prestige or headline success, opening the question of whether the digital economy is becoming socially and economically embedded or merely physically present → the state as the central actor in determining whether infrastructure becomes ecosystem depth through procurement, education, incentives, labor policy, and institutional design → benchmarking frameworks, national embeddedness measures, startup linkages, and capability ladders as the practical tools for turning infrastructure presence into long-term knowledge-economy capture → broader implications for Gulf digital policy, which must move beyond celebrating cloud and AI capacity toward deliberate oversight of value retention, talent formation, enterprise development, national participation, and durable economic intelligence.

Note: The calculations presented should be read as a structured illustrative application of our analytical framework, anchored where possible in public data and supplemented where necessary by clearly stated planning assumptions.

A Data Center Can Arrive in the GCC Long Before a Knowledge Economy Does

There is a temptation, especially in official speeches and investment headlines, to treat digital infrastructure and knowledge-economy development as nearly the same thing. A hyperscale region opens, a sovereign cloud deal is signed, a major AI campus is announced, and the assumption quickly follows that the economy has moved up the value chain. But these are not identical developments. A data center is a powerful piece of digital public and commercial infrastructure, yet it is still only one layer in a much larger system of value creation. It can improve latency, data residency, resilience, compliance, and enterprise confidence. It can also support government modernization and attract adjacent activity. What it does not do automatically is create a dense domestic ecosystem of researchers, local cloud architects, applied AI specialists, exportable software firms, or globally competitive digital products. That distinction is the starting point of this Part II.

The Gulf is not wrong to invest in this layer. In fact, the World Bank argues that digital public infrastructure, including cloud and data center capacity, has become part of the readiness journey many GCC governments and private-sector leaders are pursuing, and that Gulf states are in a position to attract and retain talent able to build, operate, and sustain large digital infrastructure investments. At the same time, the same broad digital-transformation debate also warns that productivity gains from AI do not emerge automatically at the macro level just because the technology is present; complementary innovations, diffusion, and institutional adaptation matter enormously. In other words, infrastructure can be a necessary condition for digital transformation without being a sufficient condition for a true knowledge economy.

This is exactly why the GCC needs sharper language. There is a meaningful difference between hosting digital activity and capturing knowledge-economy value from digital activity. Hosting means the region provides land, electricity, regulatory conditions, connectivity, and political stability for global firms to operate. Capturing value means the region develops local technical labor markets, domestic service providers, sectoral AI champions, applied research communities, and the ability to turn digital infrastructure into local enterprise formation and productivity growth. The first can happen relatively quickly through capital spending and policy incentives. The second is slower, more difficult, and much more dependent on education, incentives, procurement, entrepreneurship, and the structure of local demand.

That is why the opening question of this article should be posed bluntly: when a global company opens a data center in the GCC, whose knowledge economy is actually deepening? Sometimes the answer is genuinely local. Sometimes the region gains real capabilities in operations, implementation, compliance, cloud migration, and startup access. But in other cases, the data center mainly strengthens the regional delivery footprint of a foreign platform whose highest-value functions remain elsewhere. The servers are local; the control systems, product roadmaps, software layers, intellectual property, and strategic rents are not. This is not a criticism of foreign investment. It is simply an argument for more accurate diagnosis. If the GCC wants more than digital prestige, it must ask harder questions about where value accumulates and who learns from the presence of infrastructure.

The knowledge-economy issue is therefore not whether global hyperscalers should come. They clearly should, and many already have. The issue is whether their arrival is being embedded inside a broader local strategy. Without that strategy, the region may end up with excellent physical infrastructure and relatively shallow domestic capability capture. With that strategy, the same infrastructure could become a platform for local cloud adoption, AI deployment, public-sector transformation, specialist training, new firms, and deeper regional competitiveness. The point is not to dismiss the first wave of infrastructure, but to refuse to confuse it with the full story.

A data center may place world-class digital infrastructure inside the GCC, but only a wider ecosystem of firms, skills, incentives, and institutions can place real knowledge-economy value there.

Foundational distinction about what infrastructure can do and what it cannot do by itself: (i) digital infrastructure is an important enabler of a knowledge economy, but it is not the same thing as a knowledge economy; (ii) hosting capacity and capturing value are related but fundamentally different outcomes; (iii) the GCC can gain real strategic benefits from data centers without automatically generating broad domestic capability; (iv) macroeconomic and productivity gains depend on diffusion, complementary innovation, and institutional adaptation rather than infrastructure presence alone; and (v) the right opening question is not whether the Gulf is building digital assets, but whether those assets are being converted into local economic depth.

The GDP Effect Is Real, but GDP Contribution Is Not the Same as Knowledge-Economy Deepening

One of the most important mistakes in this debate is to dismiss the GDP contribution of data centers as trivial. That would be wrong. Large digital infrastructure projects do create measurable economic activity. They involve capital expenditure, construction, operations, electricity purchases, telecom services, maintenance, and professional services. They can also make local firms more productive by improving access to cloud tools, reducing latency, strengthening continuity, and enabling new digital services. Official cloud providers understandably emphasize this. AWS said its UAE region would contribute an estimated AED 41 billion to the UAE economy through 2036 and support an average of 5,984 full-time jobs annually. Google Cloud said its Doha region was expected to contribute $18.9 billion in cumulative economic activity in Qatar between 2023 and 2030 and support creation of more than 25,000 jobs in 2030 alone. Those numbers may come from commissioned impact studies, but they still highlight an important truth: digital infrastructure can have a real and sizeable economic footprint.

The problem is not that these numbers are false. The problem is that they are often interpreted too broadly. GDP contribution can come from many channels that do not, by themselves, amount to deep structural transformation. A data center can boost GDP through construction activity, imported equipment, utility purchases, and associated service spending, even if much of the strategic value chain remains controlled from outside the region. It can raise output without fundamentally changing where the most sophisticated design, software, chip, or product decisions are made. A knowledge economy, by contrast, implies that the economy is moving toward higher-value creation through skills, innovation, organizational capability, and intellectual assets. Those are related to GDP, but they are not reducible to it.

This is where the Gulf needs a more layered framework. The first layer is direct GDP effect: capex, opex, jobs, and supplier spending. The second is indirect productivity effect: what local enterprises, ministries, banks, logistics platforms, hospitals, retailers, and startups can do more efficiently because infrastructure is present. The third is capability effect: whether the region develops more engineers, implementers, cybersecurity professionals, systems architects, data specialists, and AI practitioners. The fourth is innovation effect: whether local firms create new products and services because compute, data tools, and cloud infrastructure have become more available. The fifth is strategic value-capture effect: whether the region begins owning more of the higher-order layers of digital value rather than merely consuming or hosting them. The GDP conversation is strongest when it recognizes all five layers and weakest when it collapses them into one headline number.

The UAE itself offers a useful clue about how governments are thinking about this broader picture. Official UAE strategy aims to raise the digital economy’s contribution to GDP from 9.7% in 2022 to 19.4% within 10 years. That target reflects a vision much larger than data centers alone. It suggests that the real aspiration is not merely more digital infrastructure, but a deeper restructuring of economic activity around digital sectors and capabilities. That is exactly the right ambition. But it also underlines the key point: a cloud region can contribute to that agenda without single-handedly delivering it. Infrastructure helps. It does not substitute for the wider work of economic transformation.

So the right conclusion is a balanced one. The GDP effect of data centers is real, and serious analysis should acknowledge it. But GDP contribution is not the same thing as knowledge-economy deepening, and infrastructure-related output is not identical to innovation-driven upgrading. A region can enjoy strong measured economic benefits from digital infrastructure and still remain relatively dependent on foreign platforms, imported expertise, and externally controlled value chains. That is why policymakers should stop asking only “how much GDP will this create?” and start asking “what kind of GDP, what kind of jobs, and what kind of local capability will sit behind it?”

The GDP impact of data centers can be meaningful and worth pursuing, but a rise in output does not automatically mean a rise in domestic knowledge-economy depth.

Economic layering that separates headline contribution from deeper structural change: (i) data centers can make a real contribution to GDP through construction, operations, and supporting services; (ii) those gains should not be confused with automatic progress toward a broad knowledge economy; (iii) the most important effects may come indirectly through productivity and digital adoption across other sectors; (iv) policy should distinguish between short-run output gains and long-run capability formation; and (v) the strongest question is not simply how much value is created, but how much of that value becomes locally embedded.

Hosting Infrastructure Is Easier Than Capturing the Highest-Value Layers of the Digital Economy

The central economic challenge for the GCC is not attracting infrastructure. It is capturing more of the layers above infrastructure. The basic physical layer is relatively straightforward for the Gulf to compete in. It can offer capital, land, regulatory speed, strategic location, political backing, and in some cases strong energy and connectivity assets. That makes the region attractive as a place to site cloud regions, AI compute campuses, and sovereign digital infrastructure. But the highest-value layers of the digital economy often sit elsewhere: platform architecture, core software development, model training strategy, frontier R&D, semiconductor design, global enterprise sales, and ownership of intellectual property. This means the GCC can become highly relevant to the geography of compute while still capturing only part of the wider knowledge-economy stack.

This should not be read fatalistically. It is normal for value chains to be layered. Every country does not need to do everything. The more useful question is which layers the GCC can realistically and strategically move into over time. There is a strong case for aiming beyond simple hosting and toward areas such as cloud migration services, data governance solutions, Arabic-language AI adaptation, sector-specific enterprise AI for energy, logistics, health, public administration, and cybersecurity. These are not frontier-chip layers, but they are still high-value layers. They are more locally attainable, and they connect directly to the region’s real economic structure. A Gulf knowledge economy will likely be built not by replicating Silicon Valley wholesale, but by turning regional strengths into specialized digital capabilities with export potential.

The danger comes when hosting success creates complacency. Once a country has prominent cloud providers, ministerial announcements, and international partnerships, it can become easy to assume that a knowledge economy is emerging automatically. Yet the same World Bank work on the Gulf’s digital transformation points out that countries can have strong infrastructure and policy ambition while still needing further ecosystem development, including startups, AI companies, research and development, venture capital, commercialization, and output-focused reform. Qatar’s case, for example, is described as having strong infrastructure foundations but still needing deeper ecosystem development to commercialize innovation more fully. That is not a critique unique to Qatar; it is a useful warning for the region as a whole. Infrastructure can arrive faster than innovation ecosystems mature.

This distinction between hosting and value capture also has implications for how success is measured. If success is measured only by announced investment, data center megawatts, cloud region openings, and global logos, then the GCC may look as though it has already won. If success is measured by domestic software firms, local systems integrators, exportable digital services, retention of technical talent, and the creation of regional intellectual property, then the picture becomes more demanding. That second measure is harder, but it is closer to the real knowledge-economy question. A mature strategy would embrace the first without mistaking it for the second.

The long-term test is therefore not whether the Gulf can become a place where the world’s digital infrastructure sits. It is whether the region can use that physical presence to move selectively upward in the value chain. That means identifying which capabilities need to be imported, which can be co-developed, which can be localized, and which should be deliberately nurtured as domestic strengths. The conversation becomes much more serious when the GCC stops asking only how to host infrastructure and starts asking how to build a larger economic architecture around it.

A region can become globally important in the geography of compute while still remaining only partially present in the geography of high-value digital ownership.

Value-capture logic that distinguishes strategic hosting from deeper economic control: (i) the GCC is well positioned to compete for the infrastructure layer of the digital economy; (ii) the highest-value layers often remain concentrated where software, IP, and R&D are controlled; (iii) local success will depend on moving into adjacent capabilities rather than assuming hosting is enough; (iv) infrastructure-led prestige can obscure weaker performance in commercialization and domestic innovation; and (v) the real strategic question is which layers of digital value the Gulf intends to own rather than merely accommodate.

A Practical Benchmarking Framework for Measuring Whether Hosting Is Becoming Capability

One of the most useful things Part II can do is what Part I did for electricity and infrastructure pressure: it can give the reader a way to measure the deeper issue rather than merely describe it. In Part I, the challenge was to move beyond headline investment values and convert them into megawatts, annual electricity demand, and system stress. In this Part II, the parallel challenge is to move beyond headline GDP contributions, job counts, and grand claims about the knowledge economy and ask a harder question: what exactly should be measured if we want to know whether digital infrastructure is genuinely deepening local capability? Without such a framework, public discussion will remain trapped between two weak positions. One side will say that large cloud and AI infrastructure investments must be good because they create jobs and increase GDP. The other will say that they are merely foreign-controlled enclaves that add little local value. Both positions are too crude. A proper benchmarking prevents a false binary and replaces it with disciplined judgment.

The central problem is that economic impact and knowledge-economy impact are not identical. A data center can create substantial output and still leave the most strategic layers of value elsewhere. It can generate construction demand, utility revenue, facility-management roles, and local procurement while the higher-order layers of software architecture, global product control, model ownership, advanced research, and intellectual property accumulation remain outside the GCC. That does not mean the local contribution is unimportant. It means the region needs a more nuanced way of asking what kind of value has arrived, what kind is retained, and what kind is being built over time. This is where a benchmarking framework becomes essential. It does not promise false precision. It simply helps policymakers, investors, and readers distinguish between presence, participation, and capture.

1) The first benchmark should be the Local Value Capture Ratio. Let:

L = local economic value retained in the domestic economy
T = total economic value generated by the infrastructure project and its surrounding activity

Then: Local Value Capture Ratio = L / T

This is one of the most important conceptual equations overall. It forces us to distinguish between value that merely passes through the region and value that is actually retained there. Local retained value can include wages paid to local residents, fees earned by domestic suppliers, profits retained by domestic firms, local professional services, local maintenance contracts, local digital-service spending, and other forms of embedded economic activity. Total value, by contrast, includes the entire broader project footprint, including imported equipment, foreign technical services, and externally retained profits. A project with a high gross economic footprint but a modest Local Value Capture Ratio may still be useful, but it should not be mistaken for deep local economic transformation. This ratio is the first defense against confusing size with embeddedness.

2) The second benchmark should be the Talent Localization Ratio. Let:

Ls = number of locally rooted skilled workers employed in relevant digital infrastructure and adjacent capability roles
Ts = total number of skilled workers employed in those roles

Then: Talent Localization Ratio = Ls / Ts

This ratio matters because one of the deepest questions raised by GCC digital expansion is whether the region is building a locally anchored technical workforce or mainly operating through imported expertise. Neither outcome should be moralized simplistically. Imported expertise is often necessary and beneficial, especially in early phases. But if a decade of large-scale digital infrastructure growth leaves a region still heavily dependent on externally supplied expertise in its most strategic technical roles, then the knowledge-economy outcome is thinner than official rhetoric may suggest. This ratio becomes especially revealing when measured over time. A low ratio in year one may be reasonable. A stubbornly low ratio in year ten would indicate that knowledge transfer, education, and institutional adaptation have not kept pace with infrastructure ambition. The real story is not whether foreign specialists are present, but whether their presence is helping produce deeper local capability over time.

3) The third benchmark should be the Capability Deepening Ratio. Let:

Hv = number of higher-value technical, managerial, design, and strategy roles linked to the infrastructure ecosystem
J = total jobs linked to the project and its surrounding ecosystem

Then: Capability Deepening Ratio = Hv / J

This ratio helps separate raw employment creation from knowledge-economy significance. Not all jobs do the same developmental work. Construction labor matters, but it is not the same as cloud-architecture capability. Routine operations matter, but they are not the same as advanced cybersecurity, AI deployment, platform engineering, systems design, digital governance, or enterprise solution architecture. A project can have a large employment footprint and still contribute relatively little to long-term capability if too few of the resulting roles involve higher-order learning and strategic technical development. By contrast, a smaller but more capability-intensive ecosystem may do much more to deepen the local knowledge economy. This ratio therefore encourages a richer question than “how many jobs?” It asks instead: what proportion of jobs actually build the capabilities a future digital economy depends on?

4) The fourth benchmark should be the Domestic Firm Participation Ratio. Let:

Df = value of contracts, procurement, and service opportunities captured by domestic firms
Pf = total relevant non-land, non-utility project-linked procurement and service spending

Then: Domestic Firm Participation Ratio = Df / Pf

This ratio goes straight to the heart of whether local enterprise formation is really part of the story. Are domestic firms merely peripheral suppliers, or are they increasingly involved in cloud migration, cybersecurity, enterprise integration, data governance, software customization, maintenance, digital compliance, systems consulting, and sectoral AI deployment? If local firms are weakly represented even after years of infrastructure growth, then the project may be functioning more as a foreign-operated service node than as a platform for local business capability. On the other hand, a rising Domestic Firm Participation Ratio suggests that the infrastructure is beginning to generate ecosystem depth. It indicates that local firms are not simply standing beside the investment; they are learning through it, selling into it, and building competitive capacity around it. A knowledge economy is much more likely to emerge where domestic firms are integrated into the surrounding value chain rather than confined to the edges of it.

5) The fifth benchmark should be the Innovation Spillover Ratio. Let:

Id = revenue, commercial activity, or measurable innovation output generated by local digital firms and institutions that directly leverage the infrastructure
Ei = total infrastructure-related economic activity

Then: Innovation Spillover Ratio = Id / Ei

This ratio is harder to measure cleanly, but it is conceptually indispensable. It tries to capture whether data centers and cloud regions are actually leading to local product creation, sector-specific software, AI applications, new enterprise solutions, startup growth, applied research commercialization, or other downstream innovation activity. This is where the knowledge-economy argument becomes truly serious. A region may have a strong Local Value Capture Ratio and still show weak innovation spillover if the infrastructure mainly supports mature users rather than new creators. Conversely, a moderate-sized infrastructure footprint could have outsized long-term significance if it catalyzes a disproportionately strong layer of local experimentation, commercialization, and sectoral problem-solving. The point of this ratio is not to pretend that every effect can be measured exactly in the short term. It is to insist that innovation spillover should be treated as a core benchmark rather than a rhetorical afterthought.

6) The sixth benchmark, which brings the others together, can be framed as a composite measure: the Knowledge-Economy Embeddedness Index. Let:

TLR = Talent Localization Ratio
CDR = Capability Deepening Ratio
DFPR = Domestic Firm Participation Ratio
ISR = Innovation Spillover Ratio
LVCR = Local Value Capture Ratio

Then: Knowledge-Economy Embeddedness Index (KEE) = a(TLR) + b(CDR) + c(DFPR) + d(ISR) + e(LVCR)

Where a, b, c, d, and e are weights reflecting policy priorities.

This is not meant to be a perfect econometric model. It is a strategic framework. Some countries may weight talent more heavily. Others may emphasize domestic firms, or innovation spillover, or value retention. The real point is that a project should not be celebrated as knowledge-economy infrastructure merely because it is digital, expensive, or globally branded. It should be judged against a set of metrics that reveal whether its surrounding value is becoming locally embedded, institutionally sticky, and capability-producing. A composite measure helps make that principle operational.

The framework can also be extended by introducing a Hosting-to-Embedding Progression Model. This would not be a mathematical formula so much as a staged analytical ladder:

• Stage 1—Hosting Presence: The region attracts the physical facility and captures basic infrastructure value.

• Stage 2—Service Participation: Local suppliers, service firms, and implementers begin engaging meaningfully with the project.

• Stage 3—Capability Formation: Technical labor markets deepen, higher-value roles expand, and local skill pathways become visible.

• Stage 4—Innovation Spillover: Startups, domestic enterprises, universities, and applied-research actors begin building new digital and AI uses around the infrastructure.

• Stage 5—Strategic Embeddedness: The infrastructure becomes part of a broader local economic architecture in which firms, talent, institutions, and sector specialists generate durable and increasingly self-reinforcing value.

This staged framework is useful because it prevents two equal and opposite errors. It prevents the overstatement that a data center automatically creates a knowledge economy the day it opens. And it prevents the cynical claim that infrastructure does not matter unless all value is local from the beginning. In reality, value capture is a progression. The real question is whether the GCC is moving from hosting toward embedding, or whether it is stopping at the hosting stage and calling that success. A practical application of this framework would change the way GCC policymakers evaluate digital infrastructure agreements. Instead of asking only:

• How much investment was announced?

• How many jobs will be created?

• How much GDP will this contribute?

They would also ask:

• What share of skilled roles will be localized over time?

• What proportion of surrounding service work will go to domestic firms?

• What new capabilities will local institutions acquire because of this project?

• What measurable innovation spillovers are expected within five to ten years?

• How much local value is likely to remain after accounting for imported equipment, foreign ownership, and externally retained profits?

Those questions are much harder — and much better.

This becomes especially strong when it makes clear that these ratios are not hostile to foreign investment. On the contrary, they make foreign investment more strategically useful. A region that measures embeddedness well is better able to design partnerships, education programs, procurement systems, and ecosystem policies that maximize long-term benefit. The issue is not whether global firms should lead parts of the infrastructure buildout. They inevitably will. The issue is whether the surrounding value architecture is being shaped intentionally enough to convert their presence into local capability. Benchmarking is not a technocratic exercise here; it is a way of asking what the GCC is truly trying to become.

There is also a subtle but important political-economy advantage to this framework. It helps governments avoid being trapped by simplistic publicity metrics. Announced investment is politically attractive because it is immediate and visible. Embedded capability is slower, less photogenic, and harder to communicate. But over the long run, the latter matters more. A benchmarking system that tracks localization, firm participation, higher-value employment, and spillover helps shift the policy conversation from ribbon-cutting toward accumulation. That may be one of the most valuable contributions. It gives a disciplined language for talking about what counts when the first wave of excitement fades.

This is ultimately why Part II provides such a framework. In Part I, the great analytical move was to show that infrastructure headlines should be translated into physical consequences. In Part II, the equivalent move is to show that digital-economic headlines should be translated into embedded capability consequences. Only then GCC countries can see the difference between a data center that is merely hosted in-country and a data center that is genuinely helping to build a knowledge economy.

A data center becomes part of a knowledge economy only when its value is judged not just by what it adds to output, but by what it leaves behind in local talent, firms, institutions, and capability.

What this benchmarking framework makes possible in judging whether digital infrastructure is producing real knowledge-economy depth: (i) headline GDP and job numbers become more useful when separated into direct, indirect, and embedded value; (ii) local talent formation should be measured distinctly from total employment creation; (iii) domestic firm participation is one of the clearest indicators of whether infrastructure is deepening local capability; (iv) higher-value technical and managerial roles matter more than raw job counts when assessing knowledge-economy impact; and (v) the strongest long-term benchmark is not infrastructure presence alone, but the degree to which surrounding value becomes locally retained and institutionally embedded.

In the GCC, Knowledge-Economy Benchmarks Must Distinguish Between Localization and Nationalization

One of the most important refinements Part II needs is this: in the GCC, “local” is not a precise enough category. In many other economies, when analysts ask whether the value created by digital infrastructure is staying local, they usually mean whether it remains inside the national economy through jobs, firms, tax effects, supplier linkages, and institutional capacity. In the Gulf, however, that question has always been more layered. A great deal of economic activity may be deeply embedded in-country while still being only partially embedded in the national citizen workforce. A project can generate local spending, local operations, local knowledge transfer, and local labor-market effects while still relying predominantly on expatriate talent, foreign-owned suppliers, or resident non-national professionals for many of its most sophisticated functions. That may still represent real economic development. But it does not fully satisfy a long-running GCC policy objective: the gradual nationalization of labor and capability.

This matters because the benchmark framework, if left unadjusted, could easily flatter outcomes that are only partially aligned with Gulf strategic goals. A data center might show a healthy Local Value Capture Ratio because many services are performed in-country. It might even show a rising Talent Localization Ratio if more skilled work is being done by people already based in the host economy. Yet neither of those outcomes necessarily tells us whether nationals are becoming more central to the technical, managerial, and strategic layers of the digital economy. In the GCC context, that distinction is crucial. The region’s developmental question is not only whether digital infrastructure strengthens the domestic economy in the aggregate. It is also whether that strengthening is helping to build a labor market, managerial class, and knowledge base in which citizens play a growing role over time.

That is why we need to draw a sharper analytical line between localization and nationalization. Localization means that value, activity, firms, and expertise are increasingly situated inside the country, regardless of citizenship status. Nationalization means that a rising share of strategically important roles, opportunities, and capabilities is becoming rooted in the country’s own citizen population. These two processes can overlap, but they are not identical. In fact, one can advance much faster than the other. A digital ecosystem may become more local in the sense that more work is done domestically, more resident talent is available, and more in-country firms are active, while still remaining relatively thin in terms of national participation in the highest-value technical and decision-making roles.

This is not a flaw in the benchmark framework; it is exactly why the framework now needs a GCC-specific extension. A sophisticated policy discussion in the Gulf cannot treat all in-country economic embedding as the same. It must distinguish between at least three layers of labor and capability presence.

The first layer is imported project talent. This refers to workers brought in from outside specifically to design, construct, operate, or support digital infrastructure. In the early phases of large projects, this is often inevitable and rational. The region cannot and should not expect every new hyperscale buildout or AI campus to be staffed entirely by pre-existing domestic skill pools. Imported expertise can accelerate deployment, raise standards, and transfer knowledge into the host economy. The issue is not whether imported talent is present. It will be. The issue is what institutional and labor-market effects its presence creates.

The second layer is resident non-national talent. This is especially important in GCC economies because many of them already depend on long-term expatriate professionals who are embedded in the domestic economy, understand local institutions, and often play a critical role in continuity and knowledge transfer. These workers are “local” in one economically meaningful sense: they live, work, spend, and often remain in the country over long periods. Their presence can deepen domestic capability, strengthen firms, and support ecosystem maturation. Any serious benchmark of digital-economic localization in the GCC has to count them. To ignore them would be analytically false.

The third layer is national citizen capability. This is where the policy question becomes sharper. Citizens matter not only because of labor-market preferences or political symbolism, but because nationalization is tied to long-term social contracts, public legitimacy, human-capital strategy, and the deeper goal of ensuring that economic transformation is not experienced as something happening around the national population rather than through it. If digital infrastructure expansion creates jobs, capability, and value but leaves nationals underrepresented in higher-order technical and strategic roles, then the knowledge economy may be deepening economically while remaining incomplete politically and socially. That is why the benchmark framework should now be expanded.

1) The first addition should be the National Talent Participation Ratio. Let:

Nn = number of nationals in skilled digital-infrastructure and adjacent capability roles
Ts = total number of skilled workers in those roles

Then: National Talent Participation Ratio = Nn / Ts

This ratio is the GCC-specific counterpart to the broader Talent Localization Ratio. It asks not merely how much of the skilled labor base is located in-country, but how much of it is made up of nationals. This matters because a data center ecosystem could appear increasingly embedded in-country while remaining weakly embedded in the national labor force. In the early years of expansion, a modest National Talent Participation Ratio may not be alarming. But over time, its trajectory becomes extremely revealing. If the ratio remains low or stagnant despite sustained infrastructure growth, then the digital economy is becoming more present than nationally rooted. If it rises steadily, then the region is beginning to convert infrastructure into citizen capability.

2) The second addition should be the National Capability Deepening Ratio. Let:

Nhv = number of nationals in higher-value technical, managerial, design, governance, and strategic roles
Hv = total number of higher-value roles in the ecosystem

Then: National Capability Deepening Ratio = Nhv / Hv

This ratio is far more revealing than raw national employment counts. Nationalization efforts can be overstated if success is measured only by entry-level participation or aggregate headcount. A knowledge economy is not transformed when nationals merely appear somewhere in the labor statistics. It is transformed when nationals increasingly occupy the roles that shape architecture, strategy, system design, governance, implementation leadership, commercial direction, and long-term technical evolution. A country may show some national participation in digital infrastructure while remaining highly dependent on expatriate expertise in the decisive layers of the ecosystem. This ratio helps distinguish symbolic presence from strategic participation.

3) The third addition should be the National Skills Progression Ratio. Let:

Nadv = number of nationals who progress into more advanced digital and infrastructure-related roles over a defined period
Nbase = baseline number of nationals in those roles at the beginning of that period

Then: National Skills Progression Ratio = Nadv / Nbase

This introduces something that simple stock measures often miss: movement. Nationalization is not merely about entry. It is about progression. Are nationals moving from junior technical support roles into systems architecture? From compliance work into cloud governance? From operations support into cybersecurity leadership? From vendor coordination into product management and implementation strategy? If not, then the labor market may be absorbing nationals without really deepening national capability. A healthy National Skills Progression Ratio indicates that the ecosystem is not merely hiring citizens, but helping them move upward through real technical and managerial ladders.

4) The fourth addition should be the National Retention Ratio. Let:

Nr = number of nationals retained in high-skill digital roles over a given period
Ne = number of nationals entering those roles during the same period

Then: National Retention Ratio = Nr / Ne

This is especially important in the Gulf because training and placement do not automatically translate into durable labor-market transformation. A country may produce nationals with digital certifications, place them into infrastructure-linked roles, and still fail to retain them if the incentive structures, workplace culture, advancement pathways, compensation, or job design do not support long-term technical careers. The National Retention Ratio helps expose whether national participation is becoming durable or merely episodic. Without retention, nationalization can become a revolving door rather than a developmental ladder.

5) The fifth addition should be the National Supplier Participation Ratio. Let:

Nf = value captured by nationally owned firms
Df = value captured by all domestic firms

Then: National Supplier Participation Ratio = Nf / Df

This ratio matters because labor nationalization is not the only relevant lens. Ownership and firm structure matter too. A digital ecosystem may become more domestic in the sense that more work is done by firms based in the host country, but those firms may still be foreign-owned or externally controlled. There is nothing inherently wrong with that. Yet if policymakers are serious about national economic transformation, they must also ask whether nationally owned firms are gaining meaningful footholds in the surrounding value chain. Are they winning integration contracts? Providing specialized services? Building cybersecurity practices? Developing sectoral AI tools? Participating in compliance, localization, and systems adaptation? Or are the most strategic supplier positions still concentrated in foreign or externally controlled firms? This ratio helps answer that.

These additions allow us to move from a broad embeddedness framework toward a more GCC-specific one. We can therefore define a National Embeddedness Index. Let:

NTPR = National Talent Participation Ratio
NCDR = National Capability Deepening Ratio
NSPR = National Skills Progression Ratio
NRR = National Retention Ratio
NSuPR = National Supplier Participation Ratio

Then: National Embeddedness Index (NEI) = a(NTPR) + b(NCDR) + c(NSPR) + d(NRR) + e(NSuPR)

Where a, b, c, d, and e are weights reflecting national policy priorities.

Again, the point is not pseudo-scientific precision. The point is disciplined visibility. A government may decide that national participation in high-value roles matters more than firm ownership. Another may place greater emphasis on retention or supplier development. What matters is that the framework forces the question into the open. A project should not be treated as fully aligned with national labor objectives merely because it boosts GDP, grows local spending, or even increases in-country technical employment. In the GCC, the crucial issue is whether the project is also contributing to a broader shift in who participates, who advances, and who remains in the strategic segments of the digital economy.

This refined framework also introduces a very important conceptual distinction between resident embedding and national embedding. Resident embedding means that talent, firms, and operations increasingly live inside the country’s economic fabric. This is good, valuable, and often developmentally important. National embedding means that a rising share of that same capability becomes rooted in the national population and nationally owned institutional structure. These are not mutually exclusive. In fact, the healthiest GCC digital economies will probably need both. Resident non-national professionals can deepen ecosystems, transfer knowledge, and strengthen firms. Nationals can gradually move into stronger technical and strategic positions, giving the system greater political legitimacy, social durability, and national economic depth. The mistake is to confuse one with the other. A region can become more localized without becoming more nationalized. That is precisely why the benchmarks must be adapted.

This also means that we should avoid the lazy conclusion that expatriate participation is somehow evidence of failure. That would be analytically and practically wrong. Many of the GCC’s most advanced sectors have long depended on globally mobile expertise, and digital infrastructure will be no exception. The real question is not whether expatriates are present, but whether their presence helps generate broader domestic capability or whether it simply becomes a permanent substitute for it. A smart nationalization strategy does not try to erase international talent. It tries to ensure that international talent sits inside a system of transfer, progression, and capability-building that increasingly strengthens national participation over time.

The same logic applies to firms. Foreign firms, multinational hyperscalers, and globally connected suppliers will remain central to Gulf digital infrastructure for the foreseeable future. The issue is not whether they are present, but whether their presence is surrounded by pathways for nationally owned or nationally led firms to gain capability, contracts, and sectoral depth. Nationalization in the digital economy should therefore be understood less as exclusion and more as structured deepening.

A practical way to operationalize this is to extend the earlier Hosting-to-Embedding ladder into a Localization-to-Nationalization ladder:

• Stage 1—mported Capability Dependence: Most higher-value technical and strategic roles are supplied from outside the country.

• Stage 2—Resident Localization: More technical work is performed in-country, including by long-term expatriates and resident non-national professionals.

• Stage 3—National Entry: Nationals begin entering relevant digital and infrastructure-linked professions in visible numbers.

• Stage 4—National Progression: Nationals start moving into more advanced technical, managerial, governance, and strategic roles.

• Stage 5—National Embeddedness: A meaningful share of the ecosystem’s higher-value functions becomes sustainably rooted in national talent and nationally connected firms.

This ladder is useful because it reflects the likely real-world sequence in the GCC. Most countries will not jump from imported dependence directly to full national embeddedness. There will be an intermediate period in which digital ecosystems become more local through resident talent and domestic operations before they become more national through citizen capability formation. That is normal. But it also means governments need to know where they are on the ladder, rather than assuming that any increase in local economic activity automatically fulfills long-term labor nationalization goals.

With this framework expanded, policy questions become much sharper. Governments should not ask only:

• How many jobs did this project create?

• How much local value did it generate?

• How many people were trained?

They should also ask:

• What share of strategic technical roles is now held by nationals?

• Are nationals progressing into higher-order digital functions, or concentrated in lower-value support positions?

• Are nationally owned firms increasingly participating in the value chain?

• Are citizens remaining in these careers over time?

• Is the ecosystem becoming more nationally rooted, or merely more locally operational?

These are much more demanding questions — and they are exactly the right ones for the GCC.

The real strength of this benchmarking framework is that it prevents GCC countries from settling for a generic concept of localization that might be appropriate in other regions but is too blunt for the Gulf. In the GCC, the social and political meaning of economic transformation depends in part on whether citizens experience the new digital economy as something they can enter, shape, and benefit from in durable ways. A cloud region staffed largely by in-country professionals may still be a valuable economic asset even if many of those professionals are expatriates. But a national knowledge economy requires another step beyond that: it requires that more of the advanced opportunities, firms, and decision-making roles gradually become tied to national human-capital development and nationally rooted economic life.

That is why this refinement is so important. It does not weaken the broader benchmark framework. It completes it. Part II is not merely asking whether the GCC can host digital infrastructure or even whether it can capture some local value from it. It is asking what kind of economic society is being built around that infrastructure. In the Gulf, that question cannot be answered fully without distinguishing economic localization from national capability formation.

In the GCC, a digital economy can become more local without becoming more national, which is why serious benchmarking must distinguish in-country embedding from citizen capability.

What this GCC-specific refinement makes clear about the difference between local embedding and national labor transformation: (i) a project can increase local value capture and still leave national workforce goals only partially advanced; (ii) resident non-national talent may deepen the domestic economy while not automatically fulfilling long-term nationalization objectives; (iii) the most meaningful national benchmark is not total employment, but national presence in higher-value technical and strategic roles; (iv) progression and retention matter more than entry alone when judging whether a digital sector is becoming nationally rooted; and (v) in the GCC, knowledge-economy success should be measured not only by whether value stays in-country, but by whether more of that value becomes tied over time to national talent, national firms, and durable citizen capability.

Applying the Benchmarking Framework to the Abu Dhabi AI Campus

To apply Part II to the Abu Dhabi AI campus, we should proceed in the same spirit as Part I. We keep the same physical assumptions already used there: a 5 GW full campus, PUE = 1.30, U = 0.85, and an implied midpoint investment envelope of roughly $42.35 billion based on the previously derived range. But unlike Part I, public disclosure is much weaker on local value retention, labor mix, supplier ownership, or innovation spillover than on electricity and site scale. For clarity, the benchmarking exercise below is presented for illustrative purposes rather than as a claim of final measured project outcomes. The physical assumptions carried over from Part I are anchored in public reporting and the earlier infrastructure calculations. By contrast, several of the Part II economic and labor benchmarks cannot yet be observed directly from public disclosure because the project has not published a full breakdown of local procurement, skilled-role composition, domestic firm capture, innovation spillovers, national participation, retention, or progression. Where such data are unavailable, the analysis therefore uses transparent planning assumptions and scaled reference points drawn from publicly reported regional cloud-economy studies, UAE digital-economy targets, and the broader logic of the framework itself. The purpose is not to imply false precision, but to show how the framework can be operationalized, what kinds of variables would need to be measured in practice, and how future disclosure could turn an illustrative model into a more fully evidenced benchmark.

A useful anchor already in Part II is that AWS said its UAE region involved $5 billion of planned investment, would contribute about $11 billion to UAE GDP through 2036, and support an average of nearly 5,984 full-time jobs annually. If one uses that UAE benchmark purely as a scaling device, then the Abu Dhabi AI campus, at the Part I midpoint investment, implies a gross cumulative economic footprint of roughly $93.17 billion over a similar long horizon, or about $6.21 billion per year on average, equal to about 1.1% of UAE 2024 GDP. That already tells us something important. The GDP effect could be real and large. But GDP contribution is not the same thing as knowledge-economy deepening. The benchmarks below therefore show not what has already been publicly proven, but what the project would look like under a disciplined base-case embedding scenario.

1. The Local Value Capture Ratio. Let T be total economic value generated by the project and surrounding activity. Using the scaled UAE benchmark above, let T = $93.17 billion. Because a project of this kind will still rely heavily on imported GPUs, foreign engineering, external software layers, and non-local profit capture, a reasonable base-case assumption is that about 35% of total value is actually retained inside the domestic economy. That gives:
   L = $32.61 billion
   Local Value Capture Ratio = L / T = 32.61 / 93.17 = 0.35
   This is an important result because it shows that a project can be very large in gross economic terms while still retaining only about one-third of that value locally. That is already meaningful, but it is not yet deep embeddedness.

2. The Talent Localization Ratio. Using the same AWS UAE scaling benchmark, the project implies roughly 50,684 average annual jobs across the wider ecosystem. Not all of those are skilled digital-infrastructure or adjacent capability roles, so let us assume that 35% fall into that more relevant skilled category. That gives Ts ≈ 17,740 skilled roles. If we then assume that 60% of those skilled roles are filled by locally rooted workers already based in-country, including long-term resident talent and resident expatriate specialists, then Ls ≈ 10,644. The ratio becomes:
   Talent Localization Ratio = Ls / Ts = 10,644 / 17,740 = 0.60
   This is a reasonably strong localization result in resident terms. It suggests that the ecosystem could become meaningfully in-country even while still falling short of deeper national rootedness.

3. The Capability Deepening Ratio. This is where the analysis gets more revealing, because not all jobs build the same kind of economy. If we assume that around 30% of total ecosystem jobs are genuinely higher-value technical, managerial, design, governance, and strategy roles, then:
   Hv ≈ 15,205
   J ≈ 50,684
   Capability Deepening Ratio = Hv / J = 15,205 / 50,684 = 0.30
   That means only about three jobs in ten would be doing the deeper developmental work that a true knowledge economy depends on. This is exactly the distinction Part II is trying to make. The project may have a strong employment footprint without yet generating an equally strong capability footprint.

4. The Domestic Firm Participation Ratio. Here we should isolate project-linked procurement and service spending that could realistically be contested by firms operating in the domestic economy, excluding land and basic utility sales. If we benchmark that contestable pool at roughly one-quarter of total ecosystem value, then Pf ≈ $23.29 billion. If domestic firms capture 25% of that pool, then Df ≈ $5.82 billion. The ratio becomes:
   Domestic Firm Participation Ratio = Df / Pf = 5.82 / 23.29 = 0.25
   That is not trivial. It means local firms are participating. But it also means they are still capturing only about one quarter of the contestable surrounding value chain. In other words, the ecosystem is beginning to form, but it is not yet locally dense.

5. The Innovation Spillover Ratio. This is harder to observe directly, but conceptually it is indispensable. If one assumes that local digital firms, applied-research institutions, sectoral AI developers, and implementation specialists eventually generate downstream commercial activity equal to 12% of total infrastructure-related economic activity, then:
   Id ≈ $11.18 billion
   Ei = $93.17 billion
   Innovation Spillover Ratio = Id / Ei = 11.18 / 93.17 = 0.12
   A ratio of 0.12 is not negligible. It suggests real spillover. But it also suggests that innovation remains a secondary layer rather than the dominant local outcome. That is precisely the warning built into Part II. Hosting and even service participation can expand faster than local innovation formation.

6. The Knowledge-Economy Embeddedness Index. If, purely for illustration, we weight the five core ratios equally, then:
   KEE = 0.20(LVCR) + 0.20(TLR) + 0.20(CDR) + 0.20(DFPR) + 0.20(ISR)
   KEE = 0.20(0.35) + 0.20(0.60) + 0.20(0.30) + 0.20(0.25) + 0.20(0.12) = 0.324
   On a 0 to 1 scale, that gives a Knowledge-Economy Embeddedness Index of 0.324, or 32.4 out of 100. Interpreted through the Hosting-to-Embedding progression model, that places the project somewhere between Stage 2—Service Participation and early Stage 3—Capability Formation. That is actually a very plausible Gulf result. It would mean the project is well beyond mere physical hosting, but still some distance away from full strategic embeddedness.

7. In the GCC, however, the framework has to go further because localization and nationalization are not the same thing. If we take the skilled-role benchmark above, Ts ≈ 17,740, and assume that nationals hold about 18% of those roles in the base case, then:
   Nn ≈ 3,193
   National Talent Participation Ratio = Nn / Ts = 3,193 / 17,740 = 0.18
   This immediately shows the distinction. The ecosystem may be becoming more local in a resident sense without yet becoming deeply national in a citizen-capability sense.

8. The next benchmark should be the National Capability Deepening Ratio. If nationals account for only 10% of the higher-value roles benchmarked above, then:
   Nhv ≈ 1,521
   Hv ≈ 15,205
   National Capability Deepening Ratio = Nhv / Hv = 1,521 / 15,205 = 0.10
   This is a much tougher number than the broader localization ratio, and that is exactly why it matters. It suggests that national participation could remain concentrated below the highest-value technical and strategic layers unless policy pushes much harder on progression, not just entry.

9. The UAE-specific National Skills Progression Ratio. This is a stock-to-flow benchmark rather than a snapshot. If, over a defined five-year period, the national baseline in relevant roles is about 3,200, and roughly 1,120 nationals progress upward into more advanced digital and infrastructure-linked roles during that period, then:
   National Skills Progression Ratio = Nadv / Nbase = 1,120 / 3,200 = 0.35
   A ratio of 0.35 would be respectable because it shows movement rather than mere presence. But it would also mean that most nationals in the system are still not climbing into higher-order functions fast enough to transform the overall skill structure. That is progress, but not yet deep national capability formation.

10. The UAE-specific National Retention Ratio. If we assume that 1,500 nationals enter these digital and infrastructure-linked roles during the same period, and that 975 remain in them over time, then:
    National Retention Ratio = Nr / Ne = 975 / 1,500 = 0.65
    This is not a weak result. It suggests that once nationals enter the ecosystem, a majority remain. But it is still not strong enough to guarantee long-run national deepening on its own. Retention is necessary, not sufficient. What matters is retention combined with upward progression into higher-value roles.

11. The UAE-specific National Supplier Participation Ratio. If domestic firms capture $5.82 billion of contestable project-linked value, and nationally owned firms capture only 12% of that domestic-firm share, then:
    Nf ≈ $0.70 billion
    National Supplier Participation Ratio = Nf / Df = 0.70 / 5.82 = 0.12
    This is a very revealing number because it shows how easily “domestic participation” can look stronger than genuinely national participation. The country may host firms, contracts, and service activity in-country while still leaving nationally owned firms with a relatively modest foothold in the higher-value surrounding ecosystem.

12. These UAE-specific National Embeddedness Index. Again using equal weights for illustration:
    NEI = 0.20(NTPR) + 0.20(NCDR) + 0.20(NSPR) + 0.20(NRR) + 0.20(NSuPR)
    NEI = 0.20(0.18) + 0.20(0.10) + 0.20(0.35) + 0.20(0.65) + 0.20(0.12) = 0.28
    That produces a National Embeddedness Index of 0.28, or 28 out of 100. Interpreted through the Localization-to-Nationalization ladder, that suggests the project sits much more clearly in Stage 2—Resident Localization than in Stage 5—National Embeddedness. In other words, the campus could become economically meaningful and locally operational well before it becomes deeply rooted in national talent and nationally owned capability.

What applying the Part II benchmarking framework to the Abu Dhabi AI campus shows is that: (i) the project could have a large GDP footprint and still retain only a minority share of its wider value locally; (ii) resident localization may advance much faster than genuine national participation in the GCC sense; (iii) the most capability-deepening jobs are likely to remain a minority of total employment unless policy intervenes deliberately; (iv) domestic firm participation and innovation spillover are where the difference between hosting and knowledge-economy formation becomes most visible; and (v) on a base-case reading the project looks less like a fully embedded knowledge economy and more like a powerful transition case between hosting, service participation, and partial capability formation.

The Talent Question Is the Hardest One Because Data Centers Create Jobs, but Not Always the Jobs a Knowledge Economy Needs Most

Every data center boom comes with a labor-market narrative. It will create jobs, build skills, attract engineers, and prepare the workforce for the future. Some of that is true. Construction activity creates jobs. Operating cloud regions and digital facilities creates technical roles. Regional growth in cloud services can increase demand for system administrators, cybersecurity specialists, software implementers, telecom technicians, data professionals, and enterprise consultants. Microsoft’s Kuwait announcement explicitly framed the planned Azure region as a way to advance local AI capabilities, support startups, and prepare the workforce for the future. Qatar’s Google ecosystem around the Doha region includes a Center of Excellence, training, workshops, and certification programs that QFZ says have already benefited more than 3,400 people through more than 40 programs. These are real channels through which infrastructure can influence skills formation.

But the labor-market picture is more complicated than the headline suggests. Strategy& has noted that local talent remains sparse in parts of the Gulf digital economy and that organizations often look to expatriates to plug skills gaps; it also highlighted that only 1% and 0.4% of the respective working populations of nationals in Saudi Arabia and the UAE were employed in ICT in the figures it cited, far below levels seen in the UK and US. The World Bank’s newer Gulf digital-transformation work shows strong basic and intermediate digital skills across much of the GCC and notes that Saudi Arabia and the UAE exceed 20% on advanced digital-skills availability in its referenced measures, but that still does not mean local labor markets are automatically deep enough to absorb hyperscale and AI-driven growth at every level. The tension is obvious: the region has made progress in digital readiness, yet the pool of deeply specialized, nationally rooted technical talent may still be narrower than the scale of current ambition implies.

That matters because not all “digital jobs” do the same thing for a knowledge economy. Construction jobs are not the same as operations jobs. Operations jobs are not the same as cloud-architecture jobs. Cloud-architecture jobs are not the same as AI research, product design, or domestic enterprise formation. A region can generate thousands of jobs around digital infrastructure while still relying heavily on imported expertise for the most sophisticated and strategic functions. This is not necessarily a policy failure; globally connected sectors often rely on mobile talent. But if the Gulf wants data centers to become more than enclaves of imported capability, it must care deeply about knowledge transfer, education pathways, certification, applied industry training, and the creation of local career ladders that move citizens and long-term residents into higher-value technical roles.

The most important talent question, then, is not whether expatriates will be involved. They will be, and in many cases they should be. The more serious question is whether imported talent becomes a bridge to domestic capability or a permanent substitute for it. A healthy knowledge-economy strategy would aim for the first. It would use global operators and foreign specialists not only to run infrastructure, but also to help train local engineers, deepen local enterprise use cases, support university-industry collaboration, and create more durable technical ecosystems. A weak strategy would be satisfied with world-class facilities whose most valuable functions remain externally staffed and externally controlled.

That is why the talent issue sits so close to the heart of the GCC data center boom. Knowledge economies are not built by buildings alone. They are built by people who know how to use, adapt, govern, secure, and commercialize what those buildings make possible. If the GCC wants data centers to matter economically beyond their direct footprint, the deepest multiplier may not be the rack itself, but the technical labor market that forms around it.

The long-term value of data center expansion may depend less on the number of facilities built than on whether the people needed to shape and use them become increasingly local, transferable, and economically embedded.

Talent realities that separate infrastructure employment from deeper capability formation: (i) digital infrastructure does create jobs, but not all digital jobs contribute equally to a knowledge economy; (ii) current Gulf talent progress coexists with real shortages in specialized local technical depth; (iii) imported expertise can accelerate growth, but only becomes strategically valuable if it transfers knowledge; (iv) workforce policy should focus on career ladders, not just training counts; and (v) the strongest multiplier may come from the technical ecosystem surrounding the facility rather than the facility alone.

Startups, Local Firms, and Sector Specialists Are Where the Knowledge-Economy Story Either Becomes Real or Remains Thin

If one wants to know whether digital infrastructure is becoming a genuine knowledge-economy platform, one should look less at the server hall itself and more at what local firms are doing around it. Are startups using local cloud capacity to build products? Are domestic systems integrators helping enterprises migrate and modernize? Are regional software companies developing sector-specific tools for Arabic-language workflows, energy systems, ports, hospitals, public services, logistics networks, or fintech? Are universities and applied-research centers collaborating with firms on real commercial problems? These are the spaces where infrastructure presence can begin to turn into indigenous capability.

The positive case is not imaginary. Bahrain’s cloud-first policy, launched in 2017, made cloud adoption an explicit enabler of digital transformation, and Bahrain has continued marketing itself as a regional cloud nation. Qatar’s Google partnership has linked infrastructure with training, a local office, and a Center of Excellence intended to support capability-building and startup development. Kuwait’s proposed Azure region is being framed not only as infrastructure but as part of an ecosystem for startups, research, and AI innovation. These examples suggest that GCC policymakers increasingly understand that a region is more valuable when it hosts not only compute, but also the people and organizations able to use compute well.

Still, the question is one of scale and depth. A few incubators, workshops, and startup grants do not automatically create a durable knowledge economy. Domestic firms need markets, procurement pathways, talent, risk capital, and opportunities to solve meaningful regional problems. In many Gulf states, the most promising path may not be building general-purpose consumer-tech giants, but cultivating strong sector specialists in industries the region already understands: energy, water, urban systems, ports, government services, logistics, healthcare administration, climate adaptation, mobility, and regulated enterprise environments. In those sectors, local firms can combine domain knowledge with digital tools in ways that foreign platforms may not replicate easily. This is where the GCC’s knowledge economy could become distinctive rather than derivative.

That matters because a cloud region used mainly by multinational corporations and government agencies is economically different from a cloud region that also underpins a growing layer of domestic providers and product firms. The first is useful. The second is transformative. Domestic firms are where repeated local learning happens, where young technical workers gain managerial experience, where relationships between client problems and technological solutions deepen, and where regional intellectual property can start to form. They also create a more resilient economy because they distribute knowledge beyond the walls of a few large foreign operators.

The policy implication is simple but demanding: digital infrastructure strategy should be tied explicitly to startup formation, SME capability, procurement reform, sectoral digitalization, and local commercialization. Otherwise, the Gulf may build the hosting substrate for a regional digital economy without building enough regionally rooted enterprises to inhabit it. The knowledge-economy test is therefore not whether startups are discussed alongside data centers, but whether the presence of infrastructure measurably changes the scale, sophistication, and survivability of local firms.

A data center becomes economically transformative only when local firms, not just global platforms, begin using it as a base for sustained learning, product formation, and commercial growth.

Enterprise-development conditions that determine whether infrastructure becomes local economic capability: (i) domestic firms are the clearest test of whether digital infrastructure is producing embedded knowledge value; (ii) cloud presence is most useful when it strengthens startup formation, SME upgrading, and sector-specific innovation; (iii) Gulf advantages may be strongest in applied digital solutions tied to regional industries rather than generic platform competition; (iv) incubators and training matter less than repeatable commercial pathways; and (v) the knowledge economy becomes real when local firms begin solving local and exportable problems with growing technical confidence.

There Is a Real Risk of Building an Enclave Digital Economy That Is Advanced, Visible, and Valuable but Not Deeply Rooted

One of the most under-discussed risks in the Gulf is the emergence of what might be called an enclave digital economy. By this, I mean a situation in which world-class infrastructure, major foreign providers, strong official narratives, and impressive technical assets coexist with relatively narrow local diffusion. The facilities are advanced. The logos are prestigious. The services are real. GDP contributions exist. Yet the wider economy remains less transformed than the rhetoric suggests. The highest-value jobs may remain limited in number or externally staffed. Local firms may rely heavily on foreign platforms without moving far up the value chain. Universities may train graduates who struggle to find pathways into high-end technical practice. The economy, in short, may host digital modernity without fully internalizing it.

This risk is especially plausible in capital-rich, globally connected economies. The GCC can afford to build quickly. It can import expertise. It can contract world-class partners. It can create physical infrastructure ahead of many other regions. These are strengths. But they also make it possible to substitute external supply for slower domestic capability-building. In the short term, that can be rational. In the long term, it can produce an economy that is digitally impressive but strategically thinner than it appears. The problem is not foreign presence itself. The problem is when foreign presence is allowed to stand in for local depth.

The enclave risk also has a political-economy dimension. Governments and major institutions may become highly sophisticated cloud customers while the broader base of domestic firms, educational institutions, and local technical communities evolves more slowly. That creates a two-speed digital economy. At the top, state entities, large enterprises, and foreign operators function in advanced digital environments. Beneath that, a wider national economy may adopt tools more gradually, unevenly, or superficially. A true knowledge economy requires more than elite digitalization. It requires diffusion into the middle layers of firms, professions, and institutions where productivity and skills deepen over time.

This is one reason the World Bank’s broader discussion of AI and productivity is so useful. It emphasizes that strong task-level gains do not automatically become broad macroeconomic gains unless complementary innovations and diffusion occur. That warning applies directly here. A data center can be physically present; a knowledge economy still has to diffuse socially, institutionally, and commercially. If diffusion is weak, the region may end up with a narrow digital core surrounded by a slower-moving wider economy. That would still be an achievement compared with having no infrastructure at all, but it would fall short of the transformational narrative often attached to AI and cloud investment.

The enclave problem is not inevitable. But avoiding it requires design. It means asking who gets access, who gets trained, who wins procurement, who benefits from local hosting, who builds on top of it, and how knowledge moves from global operators into domestic institutions. Once those questions are asked seriously, the debate becomes less about prestige and more about architecture. That is where we believe the GCC conversation needs to go.

The greatest long-term risk is not that the Gulf fails to build world-class digital infrastructure, but that it succeeds in building it without diffusing enough capability through the wider economy.

Structural warning signs that indicate infrastructure may be advancing faster than local economic depth: (i) a region can be digitally visible and still remain economically shallow in capability terms; (ii) fast infrastructure buildout can unintentionally substitute for slower but necessary ecosystem development; (iii) elite digitalization is not the same thing as broad-based knowledge-economy transformation; (iv) diffusion into firms, professions, and institutions is the critical missing middle; and (v) the enclave risk is avoided only when policy focuses on how knowledge travels outward from infrastructure.

The State Will Matter More Than the Market in Determining Whether Hosting Becomes Capability

This is one of those areas where the GCC should resist imported clichés. In many places, it is fashionable to say that innovation ecosystems emerge mainly through market forces once infrastructure is present. That view is incomplete even in mature economies; in the Gulf it is especially misleading. The state will matter enormously in deciding whether digital infrastructure becomes a knowledge economy. Not because governments should micromanage every outcome, but because they shape education, visas, procurement, industrial incentives, public-sector demand, research institutions, and the rules by which foreign and domestic firms interact. In the GCC context, these levers are too important to be treated as secondary.

The positive side of this is obvious. Gulf states can move quickly. Bahrain’s cloud-first framework shows how policy can accelerate adoption. The UAE’s digital-economy strategy shows how governments can set economy-wide direction. Kuwait’s partnership language with Microsoft explicitly links infrastructure to workforce preparation, startup formation, and economic diversification. Saudi Arabia’s various skills and AI initiatives point to the same ambition: infrastructure is supposed to sit inside a broader national transformation agenda rather than beside it. These are all signs that Gulf policymakers understand that the state is not a spectator in digital transformation.

But the harder question is whether states will use their leverage to demand capability formation rather than just physical presence. That means asking for more than local hosting. It means pushing for training partnerships tied to real jobs, local supplier development, applied-research collaboration, support for domestic integrators, procurement that creates pathways for smaller firms, and metrics that look beyond capex. It also means being disciplined about what should be localized and what should not. No GCC country needs to pretend it can internalize every layer of the global digital stack. But every GCC country can decide that physical presence should be accompanied by more meaningful local economic linkages than would emerge through passive hosting alone.

This is where statecraft matters more than slogans. A good digital-infrastructure policy does not just celebrate hyperscalers; it negotiates with them. It does not just approve regions; it shapes ecosystems around them. It does not just subsidize land and power; it asks what domestic skills, firms, and institutions will be stronger because the project exists. That kind of approach is neither anti-market nor protectionist. It is simply a recognition that the knowledge economy is not an incidental by-product of infrastructure. It is something that must be intentionally cultivated.

In the GCC, this will likely be done pragmatically rather than ideologically. Some countries will focus on public-sector transformation and cloud migration. Others will prioritize AI skills, startup ecosystems, free-zone experimentation, or research partnerships. The precise model may vary, but the core issue remains the same: if the state does not ask how infrastructure becomes domestic capability, market outcomes alone may not answer that question satisfactorily.

In the Gulf, the knowledge-economy value of data centers will be determined not only by where the market invests, but by what the state insists those investments must leave behind.

Governance levers that decide whether infrastructure presence turns into locally embedded economic capability: (i) GCC governments are unusually important actors in shaping digital-economic outcomes; (ii) public policy can accelerate cloud adoption, but capability capture requires more than adoption alone; (iii) the crucial shift is from approving infrastructure to bargaining for ecosystem depth; (iv) smart statecraft asks what skills, firms, and institutions grow stronger because infrastructure is present; and (v) the most successful Gulf model will be the one that uses state capacity to deepen local linkages without smothering growth.

The Long-Term Prize Is Not Simply to Host Compute, but to Build a Distinctive Gulf Model of Applied Digital Capability

The final long-term question is not whether the GCC can become a meaningful node in global compute geography. It almost certainly can. The deeper question is what it wants that position to become. One path leads to a respectable but limited outcome: the region becomes a trusted host for global platforms, a strong market for cloud services, and a user of imported AI tools adapted for local conditions. That would still matter. It would improve state capacity, enterprise productivity, and regional resilience. But it would stop short of a fuller knowledge-economy transformation.

The more ambitious path is not to out-Silicon-Valley Silicon Valley. It is to build a distinctive Gulf model of applied digital capability. That would mean using infrastructure strength, state capacity, domain knowledge, and regional demand to develop excellence in selected high-value fields: Arabic and multilingual AI deployment, public-sector digital systems, energy and utilities optimization, logistics and trade infrastructure, climate and water technologies, regulated enterprise solutions, cybersecurity, health-system digitalization, and urban operating systems. These are areas where the GCC has real context, real demand, and real reasons to want local competence. A mature knowledge economy does not need to be identical to the global frontier at every layer. It needs to be good enough, deep enough, and specialized enough to create durable value from the assets it possesses.

The World Bank’s analysis of the Gulf is again useful here because it presents digital transformation not merely as a technology agenda, but as part of economic diversification and long-term competitiveness. It notes both the region’s strong infrastructure position and the continuing need for domestic ecosystem development, commercialization, and innovation output. That combination is exactly the right lens. The point is not to deny what the GCC has already achieved. The point is to ensure that infrastructure strength becomes a base for more distinctive and durable forms of economic capability.

If the GCC gets this right, the long-term effects could be substantial. Data centers would still matter as physical infrastructure, but their true significance would lie in what forms around them: stronger technical labor markets, deeper enterprise adoption, domestic solution providers, locally relevant AI applications, exportable services, and more credible claims to economic diversification. If the GCC gets it wrong, it may still host an impressive digital future, but one whose deepest value is captured elsewhere.

That is why we end Part II where it began, but with a more precise economic lens. The long-term meaning of data centers in the GCC is not just that they bring compute closer to regional users. It is that they force the region to choose whether digital infrastructure will remain primarily a hosting layer for global firms or become the foundation for a more embedded, more local, and more strategically self-confident knowledge economy.

The real prize is not to become a place where the world’s compute sits, but a place where infrastructure, skills, firms, and sector knowledge combine into a distinctly Gulf digital economy.

Long-horizon ambition that reframes infrastructure as a base for selective but durable economic specialization: (i) hosting compute is useful, but it is not the highest expression of digital-economic success; (ii) the GCC’s best path may be specialized capability rather than imitation of every frontier layer; (iii) regional strengths in energy, logistics, government, climate, and regulated sectors can become digital strengths; (iv) data centers matter most when they anchor wider ecosystems of applied problem-solving; and (v) the long-term question is whether Gulf digital infrastructure becomes an imported service layer or a platform for a more self-confident knowledge economy.

Close Out: The Knowledge-Economy Test Begins After the Data Center Is Built, Not Before

The central lesson of this Part II is that the knowledge-economy significance of digital infrastructure cannot be read from the facility alone. A region can host cloud regions, AI campuses, and major global providers while still leaving the highest-value layers of software, research, product control, and strategic rent elsewhere. That does not make infrastructure unimportant. It makes it incomplete. The GCC is already proving that it can attract serious digital investment. The harder, more consequential question is what those investments leave behind in the domestic economy once the press releases fade.

The answer will vary by country, but the criteria are becoming clearer. The long-term payoff depends on whether data centers strengthen domestic firms, deepen cloud and AI adoption across local sectors, create meaningful technical career ladders, support local commercialization, and feed a wider economic architecture of problem-solving and innovation. If they do, then the Gulf’s digital buildout can become part of a real knowledge-economy transition. If they do not, then the region may still gain important infrastructure and service benefits, but the deeper story will remain one of hosting rather than capture. The difference between those two outcomes is not abstract. It is the difference between being a place where technology lands and a place where technology compounds into local capability.

That is why the economic debate must move beyond a single question like “Will this raise GDP?” and toward a more demanding set of questions. Whose firms grow stronger? Whose talent gets deeper? Whose institutions learn? Whose intellectual property accumulates? Whose productivity rises, and in which sectors? Those are the questions that determine whether digital infrastructure becomes a durable national asset or merely a sophisticated service layer.

The GCC’s opportunity is real. Its governments are serious, its capital is substantial, and its strategic timing is favorable. But the knowledge economy is not created by compute alone. It is created when compute is connected to people, firms, institutions, and incentives in ways that steadily thicken local economic life. The knowledge-economy test, in other words, begins after the data center is built, not before.

The real economic meaning of digital infrastructure is determined not by the opening ceremony, but by what local capabilities, firms, and institutions become stronger because it exists.

Final synthesis of what will determine whether GCC digital infrastructure becomes a true knowledge-economy asset: (i) data centers can be economically valuable without yet constituting a full knowledge economy; (ii) GDP gains, while important, must be distinguished from deeper capability and innovation gains; (iii) the real dividing line is between hosting value and embedding value; (iv) talent, firms, diffusion, and statecraft will decide the outcome more than infrastructure alone; and (v) the Gulf’s long-term success will depend on whether physical digital presence matures into local economic depth.

Taken together, Part I and Part II argue that the Gulf’s AI future will not be defined by data center scale alone, but by whether physical infrastructure, energy systems, governance, talent, and local knowledge-economy capability are built together into a durable and intelligent model.

At OHK, we help clients move beyond the surface of AI and data center expansion to understand the deeper systems that determine long-term success. Our advisory work connects digital ambition with the realities of electricity, infrastructure, cooling, governance, investment strategy, and public legitimacy so that decisions are grounded not only in technological potential, but in physical, economic, and institutional feasibility. We support clients in turning complex questions about AI, data centers, and digital infrastructure into clear strategic frameworks that link capability, location, load, resilience, regulation, and long-term value. Whether the challenge is national strategy, infrastructure planning, governance design, or investment positioning, our aim is the same: sharper judgment, stronger systems thinking, and more durable outcomes. Contact us to learn how OHK can support your next phase of AI, infrastructure, and strategic transformation.