Analysis
Are AI stocks resilient to the Iran war?
Disruption snapshot
AI stocks may no longer sell off like normal tech during war. Markets are starting to price AI infrastructure like strategic assets, closer to energy, defense, or telecom networks.
Winners: AI infrastructure firms, defense AI vendors, and companies tied to government demand. Losers: chipmakers and cloud players with high power costs, plus fragile supply chains.
Watch energy prices and power costs for AI data centers. If electricity and fuel stay high, margins and new buildouts could come under real pressure.
War usually crushes tech stocks.
When geopolitical tensions spike, investors typically dump growth companies first and move money into oil, defense, gold, and bonds. Risk rises, capital becomes cautious, and the most speculative sectors get hit hardest.
But something unusual is happening with AI stocks.
The companies building the AI ecosystem are starting to look less like traditional tech firms and more like strategic infrastructure. Think energy grids, defense systems, or satellite networks.
That distinction matters.
If markets begin treating AI infrastructure as a national capability rather than just another technology cycle, the way AI stocks behave during geopolitical crises could look very different from past tech booms.
AI is becoming strategic infrastructure
Modern AI depends on massive clusters of specialized chips, particularly GPUs used to train and run large models. Unlike earlier software waves, where most value was created digitally, AI requires enormous physical infrastructure: chips, power plants, cooling systems, and data centers.
Whoever controls that infrastructure controls the pace of AI development.
The scale of the buildout is already enormous. The International Energy Agency estimates that electricity demand from data centers could roughly double to about 945 terawatt-hours by 2030, driven largely by AI workloads. That would make data centers one of the fastest-growing sources of electricity demand in the global economy.
Governments have noticed. In the United States, policy is increasingly focused on maintaining leadership in advanced chips, data centers, and AI research. China is doing something similar through long-term industrial planning that emphasizes AI and quantum technology.
AI is no longer just a private-sector race. It has become a strategic competition between nations.
Once governments, cloud providers, and large enterprises commit tens of billions of dollars to data centers and chip capacity, those investments begin to resemble infrastructure projects rather than optional software spending. Like power plants or telecom networks, they are difficult to pause once construction begins.
War does affect AI stocks in one negative way
Training and running modern AI models requires enormous amounts of electricity. Large data centers packed with high-performance GPUs can consume hundreds of megawatts of continuous power.
The International Energy Agency estimates electricity use from data centers could rise from roughly 460 terawatt-hours today to well over 1,000 terawatt-hours by 2030 as AI adoption accelerates.
To put that in perspective, a single large AI data center can require 200 to 500 megawatts of electricity, roughly comparable to the consumption of a mid-sized city.
That creates a direct link between AI and global energy markets.
If electricity prices rise by only a few cents per kilowatt-hour, operating costs for a large data center can increase by tens of millions of dollars per year. At the scale of AI infrastructure now being planned across the United States, Europe, and Asia, energy shocks don’t just raise costs slightly. They influence where data centers are built, which companies can finance them, and how quickly capacity expands.
This is where geopolitics enters the equation.
Conflicts in energy-producing regions often drive volatility in oil, gas, and LNG markets. Those shocks ripple through electricity markets, particularly in regions already facing tight grid capacity. As AI infrastructure becomes more power-intensive, the industry becomes more sensitive to those energy disruptions.
War accelerates defense AI
Governments are rapidly integrating AI into intelligence analysis, cybersecurity, battlefield simulation, logistics planning, and autonomous systems. This shift is reversing a long-standing reluctance among many Silicon Valley companies to work with the military.
In 2025, the Pentagon’s Chief Digital and Artificial Intelligence Office announced partnerships with several major AI developers, including Anthropic, Google, OpenAI, and xAI. Each contract carries a ceiling of up to $200 million and focuses on building advanced AI workflows for national security missions.
These efforts range from cyber defense systems and intelligence analysis to logistics optimization and health-care administration for service members.
At the same time, traditional defense contractors are embedding AI into next-generation systems, from autonomous drones to command-and-control platforms.
The spending backdrop is massive.
According to the Stockholm International Peace Research Institute (SIPRI), global military expenditure reached a record $2.7 trillion in 2024, rising 9.4% in real terms from the previous year. Even a small portion of that budget shifting toward AI-enabled systems would represent tens of billions in new demand.
That creates a powerful flywheel: geopolitical tension increases defense spending, which accelerates AI adoption inside the military and intelligence sectors.
Investors are beginning to separate the AI stack into different risk profiles
Some layers remain fragile. Semiconductor supply chains, advanced manufacturing equipment, and energy-intensive data centers can all face short-term volatility if conflicts disrupt materials, energy prices, or global trade routes.
But demand for AI itself appears unusually resilient.
Governments view AI as strategic infrastructure. Corporations see it as productivity infrastructure. Both groups are racing to build capacity regardless of short-term macro conditions.
That dynamic makes the AI buildout behave more like an industrial cycle than a traditional software boom.
The capital commitments already underway are enormous. Hundreds of billions of dollars are being poured into chips, data centers, power generation, transmission infrastructure, and cooling systems. Once projects at that scale begin, they tend to continue even through periods of economic volatility.
Stopping halfway often destroys more value than finishing.
Recommended Articles
