AI's Two Economic Shocks: Why the Same Technology Can Push Inflation Up and Down

Artificial intelligence is best understood not as a single economic force but as two distinct shocks arriving on different schedules. On the demand side, AI is already driving a pronounced surge in investment and consumption, which can push prices up in the short to medium term. On the supply side, it promises a productivity-driven expansion of output that tends to lower inflation—but only over the medium and long term, and only if firms adopt the technology and workers acquire the skills to use it. That sequencing, more than any single number, is what makes AI a puzzle for central bankers and investors.

Introduction: Setting the Stage for AI's Economic Impact

The framing comes from two directions. A policy brief authored by Bank for International Settlements economists, published by SUERF in July 2024, states that AI "is poised to reshape the economy by affecting aggregate demand (via consumption and investment) and aggregate supply (through productivity)." Separately, reporting on remarks by Kevin Warsh, who was reported as Federal Reserve chairman in the coverage, describes AI as a historic supply-demand shock whose near-term demand expansion creates price and investment pressures before long-term productivity benefits spread across the broader economy.

The apparent contradiction in expert commentary—AI as both inflationary and disinflationary, both job-destroying and growth-boosting—dissolves once the two channels are separated. The demand channel works through spending today; the supply channel works through productive capacity built over years. The rest of this article unpacks each channel and then combines them into a single framework.

AI as a Demand-Side Shock: Immediate Investment and Consumption Surges

The demand-side shock is immediate. On the investment side, Warsh's remarks, as reported, point to massive private investment in AI technologies as a pronounced surge in aggregate demand—capital spending on data centers, chips, and infrastructure that registers in the economy now, not later. This is the classic pattern of a technology boom: the money spent building the new capacity arrives years before the capacity itself expands output.

Consumption adds a second, subtler channel. PwC's technical analysis of AI's macroeconomic impact traces a transmission mechanism in which AI-enhanced products increase consumer utility: greater variety through personalisation, higher quality, time saved during work and leisure, and falling prices from increased competition. As goods become more attractive relative to leisure, substitution toward consumption raises demand further, reinforced by an income effect from rising affordability. The BIS authors likewise identify consumption and wages, alongside investment, as the demand-side levers.

AI as a Supply-Side Shock: Long-Term Productivity and Efficiency Gains

The supply-side shock is a medium-to-long-term phenomenon. The BIS-authored brief is direct: by increasing productivity, AI will likely increase aggregate supply and hence output. But productivity gains are contingent—firms must adopt the technology, and workers must acquire the skills to deploy it productively.

The delay is not unique to AI. Brookings researchers, writing on AI and international trade, note that it takes time for an economy to incorporate and make effective use of new technologies, particularly complex ones with economy-wide impacts. That includes time to build a large enough capital stock for an aggregate effect, plus complementary investments in skilled people and business practices. The World Economic Forum makes the same point through historical analogy: electricity transformed economies once engineers could deploy it at scale, and the internet became transformative once digital skills spread beyond a technical elite. AI follows the same logic, with adoption trajectories that its cited data show accelerating sharply in the first years.

AI's Dual Economic Impact: Demand vs. Supply Shocks

The table below deconstructs AI's economic impact into its two causal chains, mapping each mechanism onto its macroeconomic consequences. The key analytical distinction is timing: the demand shock leads, the supply shock follows, and the inflationary or disinflationary outcome depends on which one dominates in any given window.

Read across the rows, the framework resolves the contradiction in expert commentary: the same technology produces investment-led inflationary pressure first and productivity-led disinflationary pressure later. The BIS authors state the condition explicitly—if aggregate demand trails supply, AI could have a disinflationary impact in the short to medium term; if demand leads, the impact is inflationary. Which path dominates is an empirical question, not a settled one.

Implications for Labor Markets and Inflation

For labor markets, the dual-shock model implies two distinct forces rather than one. On the demand side, the investment surge and wage effects support spending and employment in the near term. On the supply side, the BIS brief highlights automation of existing jobs and productivity-driven changes to output, with the authors flagging potential implications for inequality and fiscal sustainability alongside the inflation question. The World Economic Forum's analysis emphasises that AI's labor-market impact hinges on whether workers and firms can acquire and apply the skills that make it productive—meaning the same diffusion process that delays the supply shock also shapes who benefits from it.

For inflation, the sequencing is the whole story. A demand surge arriving before the productive capacity it builds exerts upward pressure on prices; once productivity gains diffuse through the economy, the expanded supply exerts downward pressure. The net effect over any horizon depends on the gap between the two curves, which is precisely what makes the transition difficult to manage.

Challenges for Policymakers and Investors

Warsh, as reported, framed the policy problem as one of sober judgment: the near-term demand expansion creates immediate price and investment pressures before the long-term supply and productivity benefits are fully realized. A central bank tightening against AI-driven demand risks choking off the investment that eventually delivers the supply-side payoff; easing too early risks letting demand-led inflation become entrenched. The BIS authors add that the outcome is conditional—disinflationary if demand trails supply, inflationary if it leads—so policy cannot assume either result.

Warsh also argued, according to the reporting, that AI will transform how central banks conduct monetary policy itself, moving away from what he described as "torturing data with old-fashioned models" and backward-looking statistics. For investors, the parallel challenge is uncertainty about corporate profitability and market structure while spending is front-loaded and returns are deferred. This article explains the framework, not investment advice; the magnitude and timing of these effects remain unpredictable.

Policy and Investment Considerations in an AI-Driven Economy

The decision the framework points to is this: policymakers must carefully manage the short-term inflationary pressures stemming from AI-driven demand surges while fostering conditions for long-term productivity gains to materialize—through adoption, skills acquisition, and complementary investment, since those conditions determine whether the supply shock arrives at all.

The measurable indicator to watch is threefold: central bank communications for discussions of inflation expectations, investment trends in AI infrastructure, and labor market shifts related to AI skill demand and automation rates. When central bank commentary begins weighing AI's supply-side effects alongside its demand-side pressures, that will be the observable sign the framework's long-term channel is entering policy calculations.

Sources

• Artificial intelligence, labour markets and inflation - SUERF
• Fed’s Warsh Framing AI as Historic Supply-Demand Shock and Transformation for C.Bank Data - KAOHOON INTERNATIONAL
• PWC
• The impact of artificial intelligence on international trade — Brookings
• How AI skills and experience are transforming the workplace — World Economic Forum