AI-Resilient Investing
AI-resilient investing means owning the parts of the economy that artificial intelligence cannot easily copy, automate, or make obsolete: physical, real-world businesses like mining, freight, and infrastructure. This pillar maps the theme: what it is, the research behind it, the companies that represent it, and how to invest.
- AI-resilient investing means owning businesses whose value AI cannot easily copy, automate, or make obsolete: mostly physical, real-world operations.
- Research finds physical, hands-on occupations are the least exposed to large language models, while higher-paid information work is most exposed.
- Representative areas include mining, freight and rail, out-of-home advertising, and industrial distribution; the HALO concept groups U.S.-listed names.
- It is a value and real-asset tilt: cyclical and capital-heavy, suited to long-horizon investors rather than a hiding place from drawdowns.
What is AI-resilient investing?
It starts from one question: what does a company own that an AI cannot reproduce? Software, text, and code a model can copy; a copper mine, a rail network, or a billboard on owned land it cannot. AI-resilient investing tilts a portfolio toward that second kind of business: companies whose value rests on physical assets and real-world operations rather than on work a model can automate. “Heavy asset” means a large base of property, plant, and equipment; “low obsolescence” means the asset keeps earning even as the technology around it changes.
Why does AI-Resilient Investing matter now?
The market has poured capital into anything that rides the AI wave. The mirror image of that trade is the set of businesses AI struggles to touch. The research backs the split: “GPTs are GPTs” found that about 80% of U.S. workers have at least 10% of their tasks exposed to large language models, and roughly 19% have at least half their tasks exposed, with higher-paid information work most exposed and physical, hands-on occupations least exposed (Eloundou et al., Science 2024). Owning the least-exposed end of the economy is the core of this theme.
Which market segments are most resilient to AI displacement?
The resilient areas are stubbornly physical. Freight still has to move: trucks alone carried 72.7% of U.S. domestic freight by tonnage in 2024 (American Trucking Associations). Metals still have to be mined, and the squeeze is tightening rather than easing: the IEA expects announced copper projects to fall roughly 30% short of demand by 2035 (IEA), while a single electric car uses up to four times the copper of a gasoline one (IEA, Global EV Outlook 2025) and AI data centers now pull on the same metal (S&P Global). Goldman Sachs sees the copper price climbing toward $15,000 a tonne by 2035 (Goldman Sachs Research). Even attention has a physical end: Lamar owns more than 362,000 billboards on land it controls (Lamar 10-K). None of that gets cheaper because a model got better at writing.
What makes these businesses hard to disrupt?
It comes down to replacement cost and physical work. Building a new copper mine or rail line costs far more than buying an existing one, so incumbents rarely face fresh competition. And the actual job is physical: freight has to move, copper has to be dug, billboards sit on owned land. An AI can route the trucks better or trim the back office, but it cannot do the work itself. That also tells you what the theme is in portfolio terms: a value and real-asset tilt toward cash-generating incumbents, not a growth bet. The moats here are old and unglamorous, which for durability is usually a feature rather than a flaw.
How do the AI-Resilient Investing clusters fit together?
The hub of this pillar is the Heavy Asset Low Obsolescence (HALO) concept, and its securities cover the resilient economy layer by layer. The materials layer is Southern Copper, digging the metal every electrification trend needs, and Cabot Corporation, whose specialty carbons go into tires and batteries. The movement layer spans J.B. Hunt in trucking and intermodal, Canadian National Railway on the rails, and Delta Air Lines in the air. The industrial backbone runs through W.W. Grainger, the parts distributor that keeps factories running, Cummins in engines and power systems, Lennox in heating and cooling, and Magna International in vehicle hardware. Lamar Advertising holds the out-of-home attention asset. Two glossary terms define the screen: AI displacement immunity and physical asset intensity. And for the head-to-head case, the HALO vs AI megacaps comparison sets this basket against the trade it mirrors.
Explore this theme
Which concepts explain this theme?
Which companies validate the thesis?
Which terms need disambiguation?
What should be compared side by side?
How can investors access AI-Resilient Investing?
The Heavy Asset Low Obsolescence (HALO) concept groups U.S.-listed examples, and the Roundhill HALO ETF (LOHA) tracks the Akros U.S. HALO Index of those companies.
The HALO strategy targets companies “uniquely positioned to withstand the persistent threat of AI disruption.”
— Roundhill Investments, HALO ETF (LOHA) objective
- Roundhill HALO ETF (LOHA) · Roundhill Investments Tracks the Akros U.S. HALO Index of physical-economy companies screened for AI resilience.
These references describe index-tracking relationships as a matter of fact and are not a recommendation to buy any product. Akros, as the index provider, may receive licensing fees from product sponsors. Review the product's prospectus before investing.
What are the risks of AI-Resilient Investing?
This is not a hiding place. The same physical assets that make these businesses hard to copy also make them cyclical and capital-hungry, so they can fall sharply in a downturn. Copper, freight, and advertising all track the economy, and mines, fleets, and networks need constant reinvestment that eats free cash flow when demand softens. AI resilience is a thesis, not a guarantee; automation can still cut costs and headcount inside these industries even when it never replaces the asset.
Frequently asked questions
What is AI-resilient investing?
AI-resilient investing focuses on companies whose earnings come from physical assets and real-world operations that artificial intelligence cannot easily copy, automate, or make obsolete, such as miners, railroads, billboard owners, and industrial distributors.
Which sectors are most resilient to AI disruption?
Sectors built on hands-on, physical work tend to be most resilient: mining and materials, freight and rail, energy and utilities, out-of-home advertising, and industrial distribution. Software, content, and routine information-processing businesses are the most exposed.
Is there an ETF for AI-resilient investing?
The HALO concept groups U.S.-listed examples, and the Roundhill HALO ETF (LOHA) tracks the Akros U.S. HALO Index of these companies. As always, check fees, holdings, and risk before investing.
What does the research say about which jobs AI can replace?
The 'GPTs are GPTs' study estimated that about 80% of U.S. workers have at least 10% of their tasks exposed to large language models, and roughly 19% have at least half exposed, with higher-paid, information-processing jobs most affected (Eloundou et al., Science 2024). Physical, hands-on work such as extraction, transport, installation, and maintenance scores lowest, and that is the part of the economy this theme buys.
Why does copper matter to AI-resilient investing?
Copper is a physical bottleneck AI cannot prompt away. The IEA expects announced copper projects to fall roughly 30% short of demand by 2035 (IEA), an electric car uses up to four times the copper of a gasoline one (IEA, Global EV Outlook 2025), and Goldman Sachs sees the price climbing toward $15,000 a tonne by 2035 (Goldman Sachs Research).
Sources & references
- GPTs are GPTs: Labor market impact potential of LLMs · Eloundou, Manning, Mishkin & Rock · Science, 2024-06-21
- American Trucking Trends 2025 (share of U.S. freight tonnage by mode) · American Trucking Associations, 2025-06-01
- Global Critical Minerals Outlook 2025 (copper supply gap) · International Energy Agency, 2025-05-21
- Roundhill HALO ETF (LOHA) · fund objective · Roundhill Investments, 2026-05-14
- Global EV Outlook 2025 (copper intensity of electric vehicles) · International Energy Agency, 2025-05-01
- Copper in the Age of AI · S&P Global, 2025-09-01
- Copper price forecast (long-run demand vs supply) · Goldman Sachs Research, 2026-01-15
- Lamar Advertising Company · 2025 Annual Report (Form 10-K) · Lamar Advertising Company, 2026-02-20