The New York Fed’s John Williams warned that the global supply of AI-related equipment could take more than two years to fully catch up with demand, potentially keeping upward pressure on prices. Speaking in an interview with Bloomberg, Williams linked the squeeze to rising chip prices and to higher costs for goods that rely on semiconductors.
Williams’ remarks also come alongside his view that the U.S. Federal Reserve may need to deliver “one more rate hike” this year to continue lowering inflation—while he suggested there is no clear case yet for additional tightening in 2027.
Key takeaways
- John Williams said AI equipment supply could lag demand for more than two years, a dynamic that can sustain inflation pressure.
- He attributed part of the price impact to higher chip prices, which can ripple into broader goods and services costs.
- Williams reiterated he may still support another rate hike this year, but said there’s no current reason to believe more action will be required in 2027.
- He expects investors to continue funding AI despite higher rates, viewing AI investment as less sensitive to monetary policy.
- The supply constraints could indirectly affect Bitcoin miners trying to expand into AI compute and infrastructure by tightening access to chips and power equipment.
AI buildout may outlast the supply response
Williams told Bloomberg that the AI infrastructure buildout is already affecting the cost of key inputs. In his view, supply constraints—particularly around advanced chips and related equipment—are not likely to normalize quickly.
When asked whether supply and demand for AI hardware could become more balanced over the next two years, Williams said he expects it to take longer. He added that this depends on an assumption: if AI demand continues to grow at the anticipated pace, the mismatch between supply and needs is likely to persist.
Why chip shortages matter for inflation
The Federal Reserve official framed the price channel as straightforward. As AI hardware demand rises faster than production can scale, chip prices increase. Those higher semiconductor costs then carry over into “goods that use them,” keeping inflationary pressure alive even as policymakers watch other inflation indicators.
For investors and market participants, the implication is that the inflation path may remain sensitive to real-economy bottlenecks, not only to aggregate demand. In that context, AI infrastructure delays could complicate the timing of disinflation—especially if the input shortages remain concentrated in critical components.
Fed policy: another hike this year, but not a guaranteed path forward
Alongside the AI supply commentary, Williams reaffirmed his expectation that the Fed could need to raise rates one more time this year. He stated that “it may be appropriate to have one more rate hike,” maintaining his earlier stance on the inflation-fighting process.
At the same time, he emphasized that the central question remains conditional: the policy outlook beyond the current year is not predetermined. Asked about the likelihood of additional tightening in 2027, Williams said he doesn’t currently see a reason to believe more would be needed, while also stressing that the Fed will “have to see.”
This kind of “one more” position matters because it suggests Williams views inflation risk as still active, but not necessarily as requiring a long series of additional hikes. For markets, that distinction often influences how traders price the future path of borrowing costs and risk assets.
AI investment appears resilient to tighter rates
Williams argued that strong AI investment could persist even if interest rates are higher. He characterized many AI-related investors as relatively price insensitive, saying they believe AI will produce “huge productivity gains” and “transformative changes” in the economy.
In his framing, monetary policy may not have a deep effect on AI expansion—at least not in the way it might dampen more cyclical sectors. That view highlights a potential mismatch between macro policy effects and the specific financing dynamics of technology buildouts.
Rather than expecting AI demand to fade quickly under tighter financial conditions, Williams’ comments suggest supply constraints and growth expectations may dominate the near-term picture.
Implications for Bitcoin miners pushing into AI
Beyond traditional tech markets, AI supply tightness may intersect with the crypto industry through mining firms and data-center operators exploring AI compute.
The core issue is competition for the same scarce inputs. If AI equipment shortages persist, it may become harder—and potentially more expensive—for Bitcoin miners expanding into AI-related projects to secure chips and power infrastructure needed to run compute-intensive workloads.
Cointelegraph previously reported that miners seeking AI opportunities face significant financing requirements, and that lenders have become increasingly selective. That dynamic could matter more when hardware availability is constrained, because projects may require larger upfront commitments to secure equipment and capacity.
The broader point is that even if AI investment remains resilient to higher rates, practical bottlenecks—chips, energy infrastructure, and delivery timelines—can still shape how quickly new AI compute capacity comes online. For miners making a transition into AI-adjacent business models, those constraints could influence timelines for scaling and the cost of expansion.
Williams’ comments therefore add an important macro layer to the ongoing discussion: the same supply pressures that can lift inflation may also affect which companies can turn AI ambitions into operating capacity, regardless of how rates move in the near term.
Going forward, readers should watch for signs that AI hardware supply is actually tightening or loosening—especially around chips and delivery schedules—and for any further Fed guidance that clarifies how persistent bottlenecks may alter the policy timeline.
