Warsh opens Fed AI research conference on productivity and jobs
Federal Reserve Chair Kevin Warsh opened a conference about AI, productivity and jobs in San Francisco on November 16, 2026. He brought together university researchers and business leaders to help the central bank understand how technology is changing the economy. For universities studying AI's economic effects, the event connects questions about how AI might help people create scientific ideas faster with evidence about how widely businesses are using it, its effects on jobs and its impact on productivity across the economy.
Warsh describes three possible outcomes for AI
In his published opening speech, Warsh said Federal Reserve staff had started using simplified models to study different possible situations. He stressed that these examples were meant to show what could happen, not to predict what will happen.
In the first situation, productivity improves, and both workers and business owners share the benefits. The economy keeps growing, and employment stays close to its full-employment level. In the second situation, the economy produces more and grows faster, but some workers lose their jobs and the labour market struggles to adjust. In the third situation, large investments do not produce the expected benefits. The economy slows down, unemployment rises and productivity does not improve as hoped.
A Reuters report by Ann Saphir and Howard Schneider also describes these different possible outcomes. Associated Press coverage also confirms that the remarks opened the one-day conference. The event is a public discussion to gather evidence about how AI may affect the economy. Warsh's remarks make clear that the final outcome is still uncertain.
Researchers from Stanford and MIT are on the programme
The official agenda from the San Francisco Fed names Stanford economist Nick Bloom, MIT computing dean Dan Huttenlocher, Stanford president Jonathan Levin and Brookings president Cecilia Rouse. They are taking part in a panel about how to understand the economic signs of businesses adopting AI.
The programme also includes a discussion about how companies at the leading edge of technology are using AI. Leaders from Palo Alto Networks, Stripe and Affirm will take part. Another panel focuses on putting AI into practice and includes people from Indeed, Ambi Robotics, Pearson Packaging Systems and Providence. Together, these sessions connect what the technology can do with evidence from organisations that are actually using it.
The Federal Reserve Board and the Federal Reserve Bank of San Francisco are organising the event. Their event page says the conference will help a task force study new technologies that can affect many parts of the economy. This work is one of five task forces announced in June. Federal Reserve staff support these groups, which work independently to produce careful findings and honest feedback for the Federal Open Market Committee.
Creating scientific ideas is also part of the growth question
Warsh's argument is about more than using AI to automate work that people already do. He suggests that AI could change how people create new ideas and how quickly new discoveries happen. This means the central bank needs to study how research itself becomes more productive, as well as how new tools affect jobs.
His speech refers to Are Ideas Getting Harder to Find?, a research paper published in the American Economic Review in 2020 by Nicholas Bloom, Charles Jones, John Van Reenen and Michael Webb. The authors look at two different things: how many researchers are working and how productive their research is. Their evidence from the past shows that research effort has increased while research productivity has fallen in many different areas.
The paper looks at progress in computer chips, crop yields, medical treatments and data from individual companies. These examples provide background for Warsh's question about whether AI could change how new ideas are created. They also show why measurement matters: doing more research and producing more useful results are not the same thing. A proper evaluation needs to understand how these two things are connected.
CASRAI's earlier report about the Federal Reserve's AI evidence working paper covers a separate publication. The November 16 conference provides a new place for university researchers and business leaders to share their views and help the task force with its work.
How different groups explain the issue
Warsh said the Federal Reserve has a responsibility to prepare for changes in technology and understand what they could mean for economic policy. Reuters describes his remarks as covering the possibilities between benefits shared widely across society and serious disruption. The Associated Press points to both the promise of new ideas and the risks to jobs. These reports describe possible outcomes, not results that the conference has already proved.
What university research offices can learn from the discussion
The following points are editorial suggestions for universities evaluating AI. Research offices can measure how many people use AI tools, how much time they spend on tasks, how many research results have been checked and confirmed, and how staffing changes. Keeping these measures separate gives universities a clearer picture than simply assuming that more AI use means better research productivity.
Economists studying total factor productivity can also identify exactly what a study measures. It might measure how quickly someone finishes a task, whether the same resources produce more work, or how widely a new idea or technology spreads. Each measure answers a different part of the question about economic growth raised at the conference.
For research projects that receive outside funding, a reproducible AI experiment can record how much work researchers did, which model settings they used and how much it cost to check the results. This gives research managers useful evidence when deciding which tools to use and what support to provide. It also helps them avoid confusing the results of one project with claims about economic growth across the whole country.
What happens next
The organisers say the conference will provide information for the task force's work for the Federal Open Market Committee. The task force's stated goal is to follow the evidence and produce careful findings. Researchers following this work will need to wait for the evidence and analysis the task force publishes next. They should not assume that any of the possible situations described in Warsh's speech has already happened.
Sources
- Federal Reserve Board: Kevin Warsh, Opening Remarks, November 16, 2026; checked 2026-11-17.
- Federal Reserve Bank of San Francisco: Productivity and Jobs Task Force Conference, November 16, 2026; checked 2026-11-17.
- Reuters: Warsh says emerging technologies will have significant impact, with risks, November 16, 2026; checked 2026-11-17.
- Associated Press: Warsh discusses AI's potential economic impact, November 16, 2026; checked 2026-11-17.
- Bloom, Jones, Van Reenen and Webb: Are Ideas Getting Harder to Find?, American Economic Review, April 2020; checked 2026-11-17.
