Please see below, an article from EPIC Investment Partners which discusses the potential impact of AI as a driver of disinflation. Received today – 15/05/2026
Kevin Warsh arrives at the Federal Reserve with a deceptively simple question: if companies can do more tomorrow with the same inputs they used yesterday, why should prices keep rising at all?
This is the deeper point behind his view of artificial intelligence. Warsh is not merely arguing that AI will make companies more efficient or shareholders richer. He is suggesting that it could alter the inflationary structure of the economy. If AI allows firms to automate routine services and lower the cost of starting new businesses, the Fed may need to rethink its interpretation of growth, employment, and productivity.
In theory, productivity is disinflationary. If a company produces more with the same labour and capital, unit costs fall. In a competitive market, those savings are passed to customers, slowing price growth and raising real incomes. Yet inflation persists because these gains are not automatically competed into prices. Firms with market power often retain them, while sticky costs — rents, insurance, and professional fees — remain constrained by politics, scarcity, or regulation that software cannot easily fix.
This is where AI may be different. Earlier technology cycles improved what firms could do. AI may change who can compete. For decades, organisational scale was itself a barrier to entry. Challenging an incumbent required teams of coders, lawyers, accountants, marketers, and support staff. Competition was expensive.
AI compresses that structure. A small team can now write code, generate research, answer customers, and review contracts with far fewer people. It does not eliminate expertise, but it reduces the minimum efficient scale of the firm.
That creates a potentially powerful disinflationary force. If incumbents use AI to lift margins without cutting prices, they may face leaner competitors able to undercut them. The threat is not just that existing companies become more efficient. It is that competition itself becomes cheaper.
There is a caveat. AI may create high barriers in chips, models, and data centres. These sectors are capital-intensive and may remain concentrated. But most companies will rent that infrastructure rather than own it, and the cost of using it is likely to keep falling. A small firm does not need to build a frontier model to use one, just as it did not need to build a power station to benefit from electricity.
That is the crucial distinction. The infrastructure layer may be expensive to build, but cheap to access. The disinflationary pressure comes from the industries built on top of it. If a small legal, software or financial firm can do with five people what once required 30, the incumbent’s pricing power weakens. AI does not need to make every sector perfectly competitive. It only needs to make competition cheaper.
This shift challenges the Fed’s traditional reaction function. The central bank’s inherited model assumes a familiar sequence: strong demand tightens labour markets, pushing up wages, which then feed into prices. Policy therefore leans against growth before inflation becomes embedded. A strong economy is treated as a risk to be restrained.
AI complicates that instinct. If supply is expanding through productivity, growth is not necessarily inflationary. If routine knowledge work becomes cheaper, services inflation may no longer behave as it did in a labour-heavy economy. The Fed’s question changes from “is growth too strong?” to “what kind of growth is this?”
The argument has limits. AI will not bypass planning rules for housing or solve oil supply shocks. But it may change how productivity gains pass through the economy. That is the core of the Warsh doctrine: the economy’s supply side may be changing in ways the old labour-market framework cannot see. The Fed’s task is not to assume inflation has disappeared. It is to decide whether the old signals still mean what they used to.
Please continue to check our blog content for the latest advice and planning issues from leading investment management firms.
Alex Kitteringham
15th May 2026