· 5 min read
For the first time, AI compute is poised to become a rationed national security asset. While nations have rationed fuel and electricity, compute remains uncharted. The impact of weaponising AI compute by a hostile nation, AI cartel (digital OPEC), or malicious actor, would be critical. This sets the scene for quotas being imposed, prioritising defense over non-essential uses like TikTok. The result: creating a geopolitical power struggle, not just a technological race. Governments will be compelled to protect AI and energy infrastructure with the same urgency as oil reserves or nuclear deterrence.
AI is becoming the new strategic asset, just like oil in the 20th century. As AI demand surges, access to compute could rival or surpass oil’s past strategic importance. In 1960, OPEC’s formation was underestimated; by 1973, its oil embargo triggered shortages, rationing, and economic instability. The U.S. responded with the Strategic Petroleum Reserve, the 55 MPH speed limit, and Corporate Average Fuel Economy (CAFE) standards to curb oil demand. An AI compute crisis will require similar measures, regulating both energy and AI consumption.
The stakes have never been higher. A future compute shock will not just destabilise economies—they will weaponise scarcity, with the potential to paralyse industries, energy markets, financial systems, and national security while reshaping global power. While OPEC’s oil crisis highlighted energy dependence, an AI compute shock will spotlight the need to secure digital infrastructure and the energy that powers it. Nations can no longer treat compute as an infinite resource, as was done with oil.
The fragile backbone: AI and digital infrastructure
The vulnerability of digital infrastructure is already a reality. In 2007, Estonia—an early leader in e-governance—became the first nation to face a cyberattack disrupting its critical systems. This exposed how even the most advanced nations can be crippled by digital threats.
In the U.S., the explosive growth of hyperscale data centers—critical to AI infrastructure—has raised new risks to grid stability, but the underlying theme remains energy. These facilities, especially in regions like Northern Virginia, are massive electricity consumers. During grid faults, data centers disconnect from the network in milliseconds, causing significant load losses that strain stability and increase cascading failure risks. In July 2024, data centers alone dropped 1,500 MW—equivalent to a major power plant—with reconnection taking hours. These disruptions mirror past crises, such as the 2003 blackout, prompting regulators to consider whether data centers should be subject to the same grid obligations as generators.
As AI-driven demand grows, unregulated infrastructure and electrical instability could exacerbate vulnerabilities, exposing national assets to both physical and digital threats.
The $600 billion Nvidia crash
Fast-forward to the 2020s, and the volatility that once defined energy markets has now been mirrored in the rise of AI. In 2020, OPEC shocked the world by plunging oil prices below zero, exposing the fragility of global energy markets. Today, AI is becoming inseparable from the global economy. Nvidia’s historic $600 billion loss in January 2025 was not just a dip—it was a stark warning of how the AI economy is now woven into the very fabric of financial markets. While the loss was not caused by scarcity, it served as a critical reminder: AI’s volatility has proven it can threaten global economic stability.
Now, as AI continues to permeate key sectors and industries, the potential for disruption becomes exponential. More so, if AI compute access is weaponised by adversarial nations against another nation or controlled by an AI cartel (digital OPEC), it could disrupt entire national infrastructures—just as oil once did in past geopolitical conflicts. Except the knock-on effects will be further-reaching due to AI’s permeation as mentioned above.
Unlike past energy shocks, an AI crisis will strike at the core of economies. While industries once depended on oil, they are now increasingly reliant on AI-driven analytics, automation, and intelligence. When compute power becomes scarce or weaponised, the consequences will be swift: automated supply chains will stall, AI-driven financial systems will freeze, and critical research—vital to healthcare, climate modelling, and defense—will face pronounced delays. Nations, whether at peace or in the calamity of warfare, will face heightened national security pressures, further destabilising economies and global stability.
The global digital arms race
As compute scarcity deepens, it will redraw global power lines. Nations with strong AI capabilities will be positioned to wield disproportionate power and dominance, while AI-dependent countries will find themselves digitally reliant, increasingly dependent on foreign-controlled compute access. Dynamic compute pricing will emerge, where AI services are priced based on real-time energy availability, and governments will be poised to control AI infrastructure, prioritising critical models during crises while limiting non-essential usage. This extends to everything from ChatGPT queries to non-critical industrial applications.
As AI’s energy demands escalate, nations that secure reliable AI compute and energy access can gain a significant economic and technological advantage. Meanwhile, countries that fail to adapt can expect to see industries falter under rising electricity costs and limited AI resources. entire sectors may stagnate, while those with privileged access to AI resources will thrive. Without intervention or regional regulation, AI scarcity could deepen global inequality, create new digital monopolies, and spark geopolitical conflicts over access to compute power. Nations controlling compute power will dominate intelligence, automation, and economic strength. If AI compute is weaponized by hostile nations, AI cartels (digital opec), or malicious actors, it will destabilize economies and spark a global power struggle over compute dominance. Before an AI-energy crisis unfolds, will governments act before it forces their hand? It is clear that we are on the verge of entering a new geopolitical reality where access to AI compute dictates global power.
Building the framework for AI's legislative evolution
Policymakers must prioritise AI infrastructure, invest in national resilience, and regulate compute with the same urgency as energy resources. If the energy industry can pass mission-critical legislation like Carbon Capture Storage (CCS) laws, why should AI be any different? Having contributed to South America’s first CCS legislation in Brazil, I understand the complexities of navigating such regulatory landscapes, which is why AI governance must adopt a multidisciplinary approach. This is due to AI’s growing demand for compute, which stresses energy systems and drives CO2 emissions. Legislative efforts must blend expertise in energy, CCS, low-carbon science, technology, cybersecurity, policy, and international relations. Only through collaboration can we ensure economic stability, national security, and environmental sustainability. The future is not dire, but action is paramount.
illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.
