The Geopolitics of Compute: AI and the Reindustrialization of Nations

 

The AI revolution is becoming synonymous to electricity in the 20th century, oil during industrialization and nuclear capability during the Cold War. Countries are no longer seeing AI as ‘just tech’. They are seeing it as an economic infrastructure, military leverage, geopolitical power and critical to national security. Nations globally want to build their own indigenous AI stack and systems to maintain self-sufficiency especially in the current times as deglobalization and supply chain vulnerabilities are becoming visible. Governments fear dependence on foreign AI for critical systems such as military software dependent on foreign cloud providers, national data processed externally and emergency infrastructure controlled by foreign AI systems. That creates strategic vulnerability. A country dependent on foreign AI systems risks surveillance, hidden backdoors, sanctions, cutoff from compute access, manipulation of information and economic dependence. Therefore nations like the United States, China, India, Saudi Arabia, and United Arab Emirates are investing heavily in sovereign AI stacks. Countries want to build their own sovereign clouds, local models, domestic compute, trusted semiconductor supply and national cybersecurity capability. The broader consensus is that future national power may depend not only on GDP or military size but on who controls the intelligence infrastructure underlying economics and society.

The objective of AI as national security is resurrecting the old forgotten and neglected hardware infrastructure industries that were once considered old economy sectors such as power generation, utilities, industrial cooling, construction, chip fabrication, logistics, water systems and telecom infrastructure. The AI revolution may become the largest infrastructure buildout since the industrial age. Every AI query devours enormous amounts of energy, AI models need massive data centers to run on, data centers require cooling systems, cooling systems require abundant water and energy sources, chips require advanced manufacturing and everything else depends on stable grid power. 

This blog will highlight the alarming importance of AI as a national security issue, the industries that are gaining back prominence thanks to this shift and which industries can nations focus on to build their own sovereign AI stacks and the infrastructure around it. 

Let’s look at the example of China and how they have built a country that is ready to ride the AI wave in a self-sufficient manner. 

China treated the internet as strategic infrastructure in its early days, it was a well thought out strategy and architecture. While many countries viewed the internet mainly as commerce, entertainment, or consumer technology. China viewed it as an economic infrastructure, information infrastructure and national security infrastructure. This led to a change in policy direction and the goal became to never allow core digital systems to be entirely dependent on foreign control. Examples include Huawei in telecom infrastructure, Alibaba in cloud and e-commerce, Tencent in digital ecosystems, Baidu in search and AI, ByteDance in algorithms and recommendation systems. This created domestic cloud infrastructure, local data storage, indigenous software ecosystems, internal digital supply chains ,national cybersecurity rules and internal digital governance.

Globally, the ‘Great Firewall’ was often discussed mainly as censorship. However the ‘Great Firewall’ was all about sovereignty. Strategically, it functioned as a digital sovereignty layer, domestic ecosystem protection, cyber control architecture and internal data localization mechanism. It created space for local firms to scale massively without being overwhelmed by foreign platforms early on. That helped domestic champions accumulate users, data, capital, engineering talent and AI training resources. 

Strategically, this meant that sensitive national data stayed inside national jurisdiction, domestic firms controlled major data flows and AI systems had access to enormous local datasets. Data became viewed similarly to strategic resources, industrial assets, economic infrastructure. China did not rely only on private venture capital it combined state funding, subsidies, long-term industrial planning, infrastructure financing and protected scaling environments. This allowed investments in semiconductor fabs, AI parks, cloud infrastructure, telecom equipment and smart manufacturing. 

China has increasingly integrated AI into manufacturing, logistics, surveillance, finance, military systems and industrial planning. The state viewed AI as productivity infrastructure, strategic capability, military multiplier and economic competitiveness tool. This led to national AI strategies, state-supported AI labs and compute infrastructure expansion.

China’s rise showed the world that internet infrastructure is national power, AI capability is industrial capability, compute is geopolitical leverage and digital sovereignty increasingly matters. It’s a lesson for nations to treat AI, telecom, cloud, semiconductors, and infrastructure as strategic national assets rather than only private commercial sectors.

Many regions are now pursuing their own versions of ‘digital sovereignty’:

• European Union → AI sovereignty and cloud independence

• India → semiconductor and digital public infrastructure

• Saudi Arabia and United Arab Emirates → sovereign AI infrastructure

• Japan and South Korea → semiconductor resilience

• United States → reshoring chips and compute capacity

The common realization is a nation dependent entirely on foreign digital infrastructure may eventually face strategic vulnerability. This is proving true in the current geopolitical dynamics. Besides national security dependence on outsourced digital infrastructure can seriously affect internal security ecosystem like intelligence dominance, data sovereignty, social stability, information integrity with narrative peddling by vested foreign interests.

What can other nations do to become AI savvy and become digitally sovereign nations?

AI has turned compute into a strategic resource. Previously countries imported software and this dependence was inconvenient but manageable. Now AI systems influence defence, intelligence gathering, finance, cyber operations, biotech, industrial productivity, information ecosystems. Hence making compute sovereignty is strategically important for the future. That’s why leaders are discussing sovereign clouds, national GPU reserves, domestic semiconductor capability, local AI models and protected data infrastructure. AI is becoming the defining infrastructure layer of the 21st century. AI capability now fosters GDP growth, military competitiveness, labour productivity, scientific leadership and financial power. Hence governments are subsidizing AI compute, funding semiconductor plants, building sovereign clouds, partnering with energy firms and creating national AI missions. For example European Union AI sovereignty initiatives, OpenAI partnerships with governments, Microsoft and Google building massive global AI infrastructure, sovereign AI programs in the Gulf and Asia.

The AI era is re-industrializing technology. The winners may not only be app companies or consumer internet platforms but increasingly valuable sectors will include electricity, utilities, semiconductors, industrial automation, telecom infrastructure, cooling systems, energy storage, nuclear engineering, construction engineering and optical networking. Therefore investors and governments now see AI as a convergence of software + industrial systems + energy infrastructure. If AI is the cognitive engine of national security, the physical elements such as power, semiconductors, water, and telecom represent its fuel, vital organs, and nervous system.

The power and transmission industry is gaining prominence amongst investors as AI data centres consume enormous amounts of electricity building a massive strain on national power grids. A single hyperscale AI cluster may consume as much power as a small city. This creates a huge demand for stable baseload electricity, pressure on national grids, race for cheap energy and renewed interest in nuclear, gas, solar, and geothermal. Countries with abundant reliable power gain strategic advantage. For example: United States will benefit from energy scale and tech ecosystems. United Arab Emirates and Saudi Arabia can leverage capital and energy abundance for AI hubs. The ability to compete in AI increasingly depends on whether nations can provide abundant electricity, resilient grids, and secure infrastructure. This shifts energy production from an economic issue to a geopolitical and national security issue.

AI requires stable baseload power, 24/7 uptime and high reliability. Renewables help, but AI clusters will also need gas, nuclear, storage and geothermal. Nuclear energy is increasingly viewed as a complementary solution capable of stabilizing grids while supporting the growing electricity demands of AI economies. The International Energy Agency expects nuclear to become increasingly important for AI demand later this decade. Industries seeing renewed interest include small modular reactors (SMRs) startups, reactor engineering, uranium mining as investors see nuclear as a long term AI infrastructure economy. This is why tech firms are signing nuclear agreements, financing reactors and investing in advanced energy technologies. Large technology firms are beginning to act like infrastructure and utility companies rather than pure software businesses.

Water will become a critical asset as AI data centres generate vast amounts of heat and cooling them requires water cooling systems, chilled liquid infrastructure and industrial cooling networks. Large AI data centers can consume millions of gallons of water annually, especially in warm climates. This raises concerns about sustainability, water scarcity, environmental impact and competition with local communities for resources. Desalination, water recycling, industrial water management and climate engineering will become strategically important. Cooling infrastructure will raise national energy demand as large AI facilities increase pressure on national electricity grids because cooling systems consume significant energy alongside computation itself. This can strain energy infrastructure, raise electricity prices, increase grid instability and create vulnerabilities during energy shortages. This creates a compounding energy problem where cooling systems themselves become major electricity consumers. In some data centers, cooling can account for a significant share of total operational energy usage. Governments will treat AI infrastructure planning as part of national energy-security strategy. Advanced cooling systems may emerge as a critical technology sector because efficient cooling directly affects AI operating costs, energy efficiency, infrastructure scalability and national AI competitiveness.

Telecom and Fiber Networks will gain importance as AI infrastructure depends on high-speed connectivity, ultra-low latency, global data movement and edge computing. Thus submarine cables, fiber optics, 5G/6G infrastructure, satellite internet will become strategic assets and attract investments. Despite satellites receiving more public attention, undersea cables remain the true backbone of the digital economy. More than 95% of global internet traffic travels through undersea fiber-optic cables laid across the ocean floor. Undersea cables are essential because they allow hyperscale AI infrastructure to function efficiently across continents. In the AI era, undersea cables are no longer just telecommunications infrastructure they are strategic assets that support economic power, military capability, and national digital sovereignty. 

Major cloud and AI companies are directly funding and building undersea cable systems to secure faster data transmission, reduced latency, lower operating costs, and infrastructure control. This means technology companies are evolving into global infrastructure actors with influence traditionally associated with states and telecom operators. Global defence systems, diplomatic communications, and intelligence networks rely heavily on undersea cable networks and any disruptions could affect military coordination, financial systems, intelligence operations and critical national infrastructure. Nations are competing for control over digital routes. Countries increasingly seek influence over cable construction, ownership, landing stations and regional internet infrastructure. This resembles historical competition over: shipping lanes, oil pipelines and trade routes. In the AI era, digital routes may become as strategically important as physical trade corridors. Undersea cables are vulnerable to accidental damage, espionage, sabotage, cyberattacks and geopolitical conflicts.

Semiconductor manufacturing matters as without chips, there is no AI. The strategic choke points are GPU design, chip fabrication, lithography machines, advanced packaging and rare earth minerals. That’s why export controls, chip bans, reshoring manufacturing, trusted supply chains have become geopolitical tools. The global importance of companies like NVIDIA, TSMC, and ASML reflects this. Countries are competing for semiconductor independence. Many nations are investing heavily in domestic semiconductor manufacturing to reduce dependence on foreign supply chains as advanced chip manufacturing is concentrated amongst small number of countries and companies. Governments increasingly see chip production as a strategic capability, an economic security issue and a defence priority. This has triggered major industrial policies and state-backed investments in semiconductor infrastructure. 

Countries with access to advanced semiconductor production may gain major strategic advantages in the AI era. Semiconductor shortages can now affect AI systems, automobiles, defence technology, telecommunications, cloud computing and industrial automation. This makes chips one of the most critical strategic resources of the modern economy. Since raw materials (rare earths, lithography equipment, specialised chemical) to produce chips are scattered globally and rely on a global interconnected supply chain any geopolitical tensions, trade restrictions or regional instability can impact semiconductor production and impact AI infrastructure hence having control on semiconductor production is a necessity for nations. Semiconductor leadership will shape up future geopolitical influence as countries dominating chip manufacturing will influence AI innovation, military superiority, economic competitiveness and digital sovereignty.

Real estate and industrial land will matter as AI campuses require massive land parcels, proximity to power stations, water access and logistics connectivity. This makes industrial parks, special economic zones, smart cities and energy corridors more valuable. Historically, premium real estate focused on commercial offices and urban development. In the AI era, strategic value is shifting toward industrial zones, logistics corridors, technology parks, energy-connected land and digital infrastructure hubs. Land capable of supporting hyperscale AI operations may become more economically and geopolitically important than traditional office real estate. The availability of suitable industrial land can determine a country’s ability to scale AI infrastructure, expand semiconductor manufacturing, strengthen cloud ecosystems and attract global technology investment. Industrial land will become a strategic national resource. Governments will protect and prioritize land designated for semiconductor production, energy infrastructure, data-center expansion and strategic technology projects. This could reshape urban planning, industrial policy, and national development strategies. Countries lacking sufficient infrastructure-ready land may struggle to compete in the AI economy. Tech companies are also exploring shifting compute and cloud to space due to cost and scale advantage. Thus all those nations who control space will be big winners in AI space.

Nations need to pay attention to the above industries and sectors in order to become self-sufficient in an increasingly fragmented world and protect themselves during periods of cyber warfare and instability. AI can be used as a weapon in conflict. AI weaponization will not mean killer robots but most AI conflict will occur through cyber systems, information warfare, infrastructure disruption and autonomous decision systems. Future conflict will focus less on territorial occupation and more on disabling systems, economic paralysis, hacking citizen data, information dominance and social fragmentation. Instead of bombing a city in war, nations can disrupt its enemies power systems, collapse telecom, manipulate information, disable finance and trigger panic. 

Traditional cyberattacks required humans however AI enables automated vulnerability discovery (like Anthropic’s mythos designed to find flaws and bugs), adaptive malware, real-time penetration and autonomous attack chains. An AI-driven financial attack may create panic, liquidity stress, payment outages and trust collapse, without firing a missile. In conflict health infrastructure becomes vulnerable to cyber sabotage, data poisoning, AI-generated false medical records, misinformation campaigns. During war disrupting logistics via AI could slow military mobilization, disrupt food supply, block ports and delay fuel distribution. In politics AI can create deepfake leaders, synthetic speeches, automated influence campaigns and mass disinformation. This can destabilize elections, create panic, inflame ethnic conflict and reduce trust in institutions. The dangerous part: populations may stop knowing what is real.

In conclusion AI has become a national security issue as nations and tech companies scramble to build self-sufficiency in a growingly deglobalized world. Mastering indigenous AI stacks and systems and sovereign supply chains will safeguard nations in an uncertain geopolitical environment. In addition as nations focus on building AI and digital sovereignty its gives light to investment opportunities in the old forgotten sectors and industries.

(Disclaimer: The opinions expressed within this article are personal opinions of the author. The facts and opinions appearing in the article are views of the author in general and the author does not hold any legal responsibility or liability for the same.)