Federal Reserve’s Daly cites regulatory barriers to AI-driven productivity growth

San Francisco Federal Reserve President Mary C. Daly spoke at an event called “The AI Moment? Possibilities, Productivity, and Policy” in San Jose, California, acknowledging AI’s potential but flagging a familiar problem: regulation is getting in the way of the technology actually delivering sustained productivity growth.

Despite billions pouring into AI infrastructure and development, the macroeconomic evidence for meaningful productivity gains remains limited. Most studies Daly referenced indicate only modest advancements in aggregate productivity so far.

The electricity analogy, again

Daly drew parallels between the current AI moment and previous technological revolutions, specifically the adoption of electricity and information technology. She pointed to modest productivity impacts already visible in specific sectors like call centers, software development, and financial services.

To get a better read on what’s actually happening, Daly argued the Fed needs to dig into disaggregated micro data and ramp up business outreach. She compared this approach to former Fed Chairman Alan Greenspan’s strategy during the 1990s computing boom, when Greenspan relied on granular business intelligence to understand productivity trends that weren’t yet showing up in official numbers.

Regulation as the bottleneck

Daly specifically highlighted state-level AI regulations as a potential drag on innovation, arguing that these rules could disproportionately burden startups compared to larger companies.

Daly emphasized that sustained productivity growth from AI depends on more than just adopting the technology itself. Companies need to fundamentally reorganize their work processes and organizational structures to capture the full benefit.

What this means for investors

Daly’s remarks carry real weight for anyone with exposure to AI-related assets. The Fed’s assessment of AI’s productivity impact feeds directly into its outlook on economic growth, inflation, and interest rates. If the Fed concludes that AI isn’t yet delivering meaningful productivity gains, that removes one potential argument for a more optimistic growth outlook.

The regulatory dimension matters too. If state-level AI rules continue to proliferate without federal preemption, decentralized AI projects could find themselves navigating a patchwork compliance landscape that Daly described as making innovation more expensive and unpredictable, with the burden falling hardest on startups least equipped to absorb it.

If the Fed starts surfacing micro-level evidence of AI productivity gains through its business outreach channels, that could shift the narrative faster than the aggregate statistics. The signal to monitor isn’t the headline GDP number. It’s the Fed’s own research publications and regional survey data, where early signs of a productivity inflection would appear first.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.