Senators press Powell on BLS staffing, FedNow costs and stablecoin demand; Fed flags monitoring work on AI

Several senators used the hearing to press Federal Reserve Chair Jerome H. Powell on a range of data, payment system and technology issues, including staffing cuts at the Bureau of Labor Statistics, the Fed’s FedNow instant payment service, and regulatory treatment of stablecoins.

BLS and data quality

Senators warned that reductions in Bureau of Labor Statistics (BLS) staffing have increased the share of inputs the agency must estimate when compiling the Consumer Price Index and other measures. Powell said the Fed is ‘‘concerned’’ about any deterioration in public data and that accurate inputs are important to both public‑ and private‑sector decision‑making. He said he was not aware of a specific BLS study the Fed could no longer replicate but welcomed further committee information if specific gaps exist.

FedNow rollout and funding

Several senators asked about FedNow, the Fed’s real‑time payments system, and its implementation costs. Senators cited an initial program cost of roughly $545 million with additional spending of about $245 million expected in 2025. Powell said the Fed is required to run FedNow on a self‑funding basis and that uptake has been steady but slower than some expected; he said the Fee structure is intended to cover direct and indirect costs but that he could not guarantee the target would be met immediately.

Stablecoins and demand for short‑term Treasuries

Senators discussed the recently passed Senate bill (referred to in the hearing as the Genius Act) that would create a regulatory framework for payment stablecoins backed 1:1 by cash or short‑dated Treasuries. Powell said such a framework is ‘‘a great thing’’ and that increased demand for Treasury bills from stablecoin backing could put downward pressure on short‑term yields, though he did not quantify the effect.

AI, credit models and data uses

On artificial intelligence, Powell said the Fed has research underway and monitors global developments. He noted both the potential for AI to raise productivity and the uncertain timing and scale of effects. Committee members raised concerns about AI‑driven credit models and fair‑lending risks; Powell said the Fed is aware of model bias risks and engages with industry and researchers to mitigate them.

Why it matters: Reliable price and labor‑market data, the structure of payment systems and the design of regulatory frameworks for stablecoins and AI models influence how well monetary policy and financial supervision can function. Senators asked for more information and documentation where they saw potential risks to policymaking and market functioning.