New York Federal Reserve President John Williams says the supply of equipment needed for the artificial intelligence buildout could take more than two years to catch up with demand, extending a bottleneck that is already contributing to higher prices for technology inputs. Williams increasingly sees AI investment as a near-term source of inflationary pressure, even as he expects the technology to improve productivity over a longer horizon.
In an October 13 interview reported by Reuters, Williams reaffirmed that one additional Federal Reserve rate increase could still be appropriate in 2026. Asked separately about the AI supply imbalance, he said it could take longer than two years for supply to catch demand if the current investment impulse persists. The timeline is conditional rather than a fixed Fed forecast, because both future demand and new productive capacity remain uncertain.
Williams had already identified the imbalance in a July 15 New York Fed speech, where he said demand was outstripping supply for semiconductors, power transformers and other equipment needed for AI infrastructure. Those components also feed into products used elsewhere in the economy, allowing higher technology costs to spread beyond data centers themselves.
AI Demand Adds Another Inflation Constraint
The issue has become more prominent as the year progressed. Williams identified robust demand associated with AI investment alongside energy disruptions as continuing sources of inflation pressure. He said the inflationary effect of the AI demand shock had become increasingly salient, although he also stressed that there was no evidence these shocks had yet developed into broader persistent inflation.
That distinction matters for monetary policy. Williams currently expects inflation to move substantially lower in 2027 as tariff effects fade and energy pressures normalize, while underlying inflation indicators have also moderated. His base case therefore remains disinflation rather than an open-ended tightening cycle, despite the possibility of one additional rate increase this year.
For crypto markets, that creates a more nuanced macro backdrop than a simple assumption that AI inflation automatically means higher rates indefinitely. Bitcoin’s relationship with monetary policy has recently become less mechanically tied to the Federal Reserve, while higher Treasury yields can affect crypto companies with capital-intensive balance sheets more directly through financing and valuation channels. The monetary-policy impact is therefore likely to differ across crypto assets and businesses rather than transmit uniformly across the sector.
AI Infrastructure Pressure Extends Into Crypto
The AI investment boom also intersects with crypto through physical infrastructure. Bitcoin miners and data-center operators increasingly compete for electricity, sites, transformers and computing equipment as some mining companies redirect capacity toward AI and high-performance computing. That competition means the AI supply squeeze can affect crypto companies even without transmitting first through interest rates.
The overlap is already visible in the industry’s shift toward AI hosting, a trend that has raised questions about whether artificial intelligence is competing with crypto for capital and infrastructure. It does not mean every dollar spent on AI would otherwise have entered crypto, but both sectors draw on some of the same power and data-center resources.
Williams’ latest comments therefore point to two separate transmission channels. Persistent AI demand can influence inflation and monetary policy while simultaneously tightening access to infrastructure used by parts of the crypto industry. How long either effect lasts will depend on whether equipment supply expands quickly enough to close the gap without a comparable slowdown in AI investment.
