San Francisco Fed President Mary Daly said Tuesday that the Federal Reserve needs to carefully analyze available data to establish whether artificial intelligence is producing measurable productivity gains that could permit faster economic expansion without generating inflation.
Daly made the remarks in prepared comments delivered at San Jose State University during an event hosted by the Silicon Valley Leadership Group. Her comments came amid public claims by the Trump administration that AI is already lifting productivity and predictions by some economists that escalating AI investment will spur growth without igniting inflation, in a manner likened to the 1990s computer-led productivity surge.
"Most macro-studies of productivity growth find limited evidence of a significant AI effect," Daly said.
She offered two possible interpretations for the current state of evidence. One is that it remains premature to detect the impact of company-level AI investments across some industries, where benefits may take time to materialize. The other is that broader, economy-wide transformations attributable to AI simply have not yet occurred.
Daly emphasized the need for careful, data-driven assessment before concluding that AI is delivering sweeping productivity improvements. Her comments underline the Federal Reserve's interest in understanding whether technological adoption can alter the trade-offs between faster growth and the risk of higher inflation.
Summary
Mary Daly called for a thorough examination of data to determine whether AI is driving productivity growth and supporting quicker economic expansion without producing inflationary pressures. She noted limited support for a significant AI effect in macroeconomic productivity studies and suggested the absence of evidence could reflect timing or the lack of a broad-based transformation so far.
Key points
- The Fed needs detailed data analysis to assess AI's effect on productivity and inflation dynamics.
- Government officials and some economists argue AI could boost growth without raising inflation, drawing parallels with past technology cycles.
- Existing macro studies generally show limited evidence that AI has produced a significant productivity impact to date.
Risks and uncertainties
- Insufficient current data - It may be too early to detect industry-specific productivity gains from company AI investments.
- Limited macro evidence - Studies to date do not strongly support a widespread AI-driven productivity increase, leaving outcomes uncertain for economic growth and inflation.
- Timing of transformations - Economy-wide effects, if they occur, could require more time to appear, complicating near-term policy assessment.