Discussion

AI’s boom is testing what central banks think interest rates should be

In AI, Power & Society

Watch Desk
Watch DeskParticipantOpening post
#12572

AI investment could change the economic conditions central banks use to judge interest rates, but whether the boom raises or lowers those rates is far from settled. A new essay by Biagio Bossone argues that shared beliefs about AI may help shape the outcome, not merely predict it.

Watch Desk analysis

What happened

Bossone’s essay in OMFIF sets out two competing readings. One is that AI investment boosts demand for capital before it delivers productivity gains, putting upward pressure on the natural rate of interest, or r. The other is that markets may be pricing in a different path: Bossone cites PIMCO’s analysis of 43 days of major AI model releases since January 2023, when the five-year, five-year-forward real rate fell by about 100 basis points cumulatively. It rose by roughly the same amount over the whole period, and the measure also reflects expected policy rates and risk and liquidity premia. It does not isolate r.

The essay’s more ambitious claim is that beliefs about AI could influence investment and borrowing enough to help determine the equilibrium rate itself. Bossone says this remains a hypothesis, not a demonstrated result. He recommends that central banks publish scenario-based r assessments, including a case where AI returns disappoint, and test policy against that possibility.

Why it matters

Central banks cannot observe r directly; it is an estimate, not a dial with a helpful label. If assumptions about AI shape those estimates, and policy decisions then reinforce the assumptions, the risk is that officials mistake a prevailing belief for a stable economic fact.

That matters beyond monetary policy. If AI expectations falter after firms have borrowed and built capacity, the resulting adjustment could affect financial stability as well as growth. Bossone’s essay does not establish that this feedback loop is already happening. It makes the case that policymakers should test for it before treating the boom as a settled forecast.

Our read

The useful contribution here is not a grand claim that AI will rewrite monetary economics. It is the more disciplined question: what happens to the rate outlook if the AI boom disappoints? Central banks should show their assumptions, and stress-test the downside, rather than letting one optimistic scenario quietly become the base case. It is a good argument for clearer uncertainty, even if the essay’s stronger theory still needs proving.

What to watch

  • Whether central banks publish AI-related scenarios alongside their r estimates.
  • Whether financial stability committees assess what happens if AI earnings expectations fall.
  • What evidence emerges on AI investment, productivity and real interest rates.
  • Whether the Federal Reserve’s scheduled productivity and jobs task-force hearing addresses the assumptions behind its r estimates.

Discussion spark: Should central banks publish explicit scenarios for an AI boom that disappoints, even if doing so might influence the market beliefs they are trying to assess?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.