Throughout economic history, technological innovation has consistently acted as a deflationary force, pushing down prices and increasing accessibility to goods and services. Artificial intelligence represents the latest chapter in this ongoing narrative, but with a distinctive twist. While AI promises to dramatically reduce costs across service industries, it simultaneously drives increased demand for the raw materials necessary to build and power AI infrastructure. This unique combination of deflationary and inflationary pressures could reshape economic landscapes in profound and sometimes contradictory ways.
The Deflationary Power of AI in Service Industries
Services represent the dominant sector in advanced economies, accounting for approximately 70% of consumer spending in countries like the United States. This broad category encompasses everything from healthcare and education to banking, insurance, and customer service operations. Historically, these sectors have been notably resistant to significant cost reductions due to their labor-intensive nature and the specialized skills, education, and credentials required by workers.
Artificial intelligence is now challenging this established dynamic by automating increasingly complex cognitive tasks that previously required human intervention. The impact is already visible across various service sectors, beginning with relatively straightforward applications but rapidly expanding into more sophisticated domains.
Customer service represents one of the earliest and most visible areas of AI implementation. AI-powered chatbots and virtual assistants are now capable of handling a substantial portion of routine customer inquiries, complaints, and service requests without human involvement. For many businesses, call centers represent a significant operational expense, typically accounting for 2-3% of overall operating costs—and even higher percentages in complex industries like telecommunications. By implementing AI solutions in these environments, companies can achieve substantial cost savings while maintaining or even improving service levels.
The real-world impact of these implementations is already demonstrable. Klarna, the Swedish fintech company, recently deployed an AI assistant powered by OpenAI technology that effectively handled the workload equivalent to 700 full-time customer service employees. This AI system now manages approximately two-thirds of Klarna’s customer service conversations, resulting in significant operational cost reductions. While every business situation has unique characteristics, even partial implementation of similar solutions across industries could substantially reduce consumer-facing service costs in competitive markets.
Perhaps more surprisingly, AI’s deflationary potential extends beyond routine tasks into highly complex professional services. Fields like healthcare, legal services, and financial advisory have traditionally commanded premium pricing partly because practitioners require extensive education and training—investments that must be recouped through higher service fees. Healthcare costs in particular represent a major financial burden for both individuals and governments, with medical expenses being the leading cause of personal bankruptcy in many countries.
AI systems are beginning to transform these dynamics by augmenting professional capabilities in meaningful ways. In healthcare, AI-enabled diagnostic tools can rapidly analyze medical data, identify patterns, and assist physicians in making more accurate diagnoses. This not only increases efficiency but can significantly reduce costs by minimizing diagnostic errors and preventing unnecessary treatments. Similar trends are emerging in legal services, where AI tools can perform document review and legal research tasks that previously required substantial billable hours from junior attorneys.
The Inflationary Countertrend: AI’s Material Requirements
While AI demonstrates remarkable efficiency in reducing service costs, the technology still relies on substantial physical infrastructure that requires significant amounts of hardware, energy, and raw materials. This material dependence creates an opposing inflationary pressure on certain commodity categories.
At the most fundamental level, AI systems—particularly large language models and other advanced applications—require vast computational resources to operate. The training phase for these models demands even greater computing power, with requirements growing exponentially as models increase in complexity and capability. This computational intensity translates directly into increased energy consumption.
The Electric Power Research Institute recently revised its projections for data center power consumption upward after incorporating more realistic AI growth scenarios. Their latest estimates suggest that data centers could consume over 9% of total US electrical power output in the coming years—a substantial increase from previous forecasts. These projections continue to rise as analysts adjust to the unprecedented growth rate of this new demand source.
Beyond energy requirements, AI infrastructure development is driving demand for specific metals and materials essential to computing hardware and data center construction. Copper, a crucial component in electrical wiring, circuit boards, and cooling systems, faces particularly strong demand pressure. Nations with historically strategic approaches to commodity markets, such as China, have already begun accumulating copper stockpiles, which have reached multi-year highs in anticipation of sustained demand growth.
The recent demonstrations of AI-powered robots by companies like Figure and Tesla further illustrate how AI’s material requirements may expand. As these robots progress from laboratory demonstrations to commercial deployment, they will create additional demand for the same critical materials required by existing AI infrastructure. In essence, we may be witnessing the early stages of an entirely new industry with material requirements comparable to the automotive sector at scale. While AI reduces the labor costs associated with operating and maintaining these systems, it cannot eliminate the fundamental need for the commodities that make them function.
The Bifurcation of Inflation: Services vs. Commodities
This contrasting dynamic between AI’s service efficiency and material requirements points toward a likely bifurcation in inflation trends across different economic sectors. Consumers may simultaneously experience declining costs in service categories like customer support, healthcare, education, and financial services due to AI-driven efficiencies, while facing rising prices for commodities, energy, and goods with high material content.
In the service domain, AI implementation can reduce costs in multiple ways. By automating routine tasks, companies can serve more customers with fewer human employees. AI systems can operate continuously without breaks, sick days, or shift limitations. They can often provide more consistent service quality while eliminating human error in routine processes. As these efficiencies compound and competition forces companies to pass savings to consumers, service costs should theoretically decline.
Simultaneously, the massive infrastructure requirements for AI will likely continue driving demand for metals like copper, specialized semiconductor materials, and various rare earth elements necessary for advanced computing components. Energy resources such as natural gas and uranium may also see increased demand as power requirements for AI data centers grow. These demand pressures, coupled with the inherent constraints in rapidly expanding commodity production, create the conditions for potential price increases.
The energy dimension of this equation deserves particular attention. AI systems require not just significant quantities of energy but also highly reliable power sources with minimal interruption risk. This requirement typically translates to increased demand for baseload power generation sources like natural gas and nuclear power, rather than intermittent renewable sources. The Electric Power Research Institute’s revised projections highlight how AI power demand estimates continue climbing as analysts adjust to growth rates that exceed traditional forecasting models.
Implications for Economic Policy and Investment
This potential bifurcation in inflation patterns presents novel challenges for economic policymakers and investors. Central banks traditionally manage inflation through monetary policy tools that affect the economy broadly, without the precision to address divergent inflation trends across sectors. If AI simultaneously creates deflationary pressure in services and inflationary pressure in commodities, conventional monetary policy approaches may struggle to maintain balance.
For investors, this divergence creates both risks and opportunities. Service-focused businesses that effectively implement AI may benefit from reduced operational costs while maintaining pricing power, potentially increasing profit margins. Commodity producers and companies involved in critical AI infrastructure development may benefit from increased demand and rising prices for their outputs. Conversely, businesses caught between these trends—requiring substantial material inputs while facing AI-driven competition in their service offerings—could face margin compression.
From a broader economic perspective, this bifurcation may also influence labor markets and income distribution. As AI reduces labor requirements in service industries, workers may face transition challenges as employment shifts toward areas less susceptible to automation or toward the development and maintenance of AI systems themselves. The net impact on employment and wages will depend heavily on how quickly workers can adapt to changing skill requirements and how effectively educational systems can prepare people for the evolving job market.
Conclusion
Artificial intelligence represents a technological force with the potential to reshape economic patterns in complex and sometimes contradictory ways. Its remarkable efficiency in service provision promises to reduce costs across industries that have historically been resistant to productivity improvements. Simultaneously, its material and energy requirements create substantial new demand for commodities and power generation.
This dual impact—deflationary for services but inflationary for commodities—creates a novel economic dynamic that defies simple categorization as either inflationary or deflationary in aggregate. As AI continues to develop and deploy across industries, policymakers, business leaders, and investors will need to carefully monitor these divergent trends and adapt their strategies accordingly.
While the broader societal impacts of artificial intelligence extend far beyond these economic considerations, understanding this bifurcation in inflation patterns provides an essential foundation for navigating the complex economic landscape that AI is helping to create. The winners and losers in this new economic environment will likely be determined by how effectively organizations anticipate and adapt to these dual pressures of service deflation and commodity inflation.
