Why did Google just hire a Nobel laureate to study AI's effect on jobs?

The four additions and what each one brings

Google announced the expansion on September 18, 2026. The program now includes two new academic advisors and two new directors who will run empirical research day to day.

Philippe Aghion joins as an Academic Advisor. He won the Nobel Prize in Economics in 2025 and holds chairs at INSEAD and the Collège de France. His work on innovation-led growth and creative destruction will be used to model AI's long-term macroeconomic trajectory. He sits alongside Nobel laureate Michael Spence and Cambridge professor Dame Diane Coyle on the advisory group.

Ajay Agrawal becomes a Visiting Fellow. He holds the Geoffrey Taber Chair in Entrepreneurship and Innovation at the University of Toronto's Rotman School of Management. Agrawal will collaborate directly with David Autor, who already serves as a Fellow and heads MIT's Department of Economics. Their joint focus covers the economics of AI and scientific discovery, AI and robotics, and how AI can expand the frontier of human welfare.

Anu Madgavkar and Daniel Rock step in as Directors of the AI & Economy Research Program. Madgavkar spent two decades at the McKinsey Global Institute, most recently as a Partner. She has advised governments, multilateral institutions, and business leaders worldwide, contributing regularly to the World Economic Forum and the United Nations. At Google she will lead empirical research on global AI diffusion, small business ecosystems, and the workforce impacts of generative AI.

Rock joins from the Wharton School at the University of Pennsylvania. He is an AI2050 Early Career Fellow and an MIT Digital Fellow. His past work includes foundational papers on technological transitions. At Google he will bridge frontier model telemetry with rigorous econometrics to analyze enterprise productivity, labor restructuring, and scientific discovery.

The four new appointees join Alex Imas, Director of AGI Economics at Google DeepMind, and Zanna Iscenko, AI & Economy Lead in the Chief Economist's Office. Together the six-person leadership team steers the research agenda.

What the program actually measures

The expansion did not happen in a vacuum. Google recently launched the AI & Economy ATLAS v1.0, an interactive open-access site that tracks how people use Google's AI tools at work and in daily life. Adoption patterns are only the starting point.

Measurement happens in real time across four core areas. Future of work. Productivity and growth. Global technology diffusion. AI's impact on scientific discovery. Each area requires fine-grained data paired with rigorous economic inquiry.

Small business ecosystems draw Madgavkar's focus because most employment sits outside large enterprises. Rock's telemetry-to-econometrics pipeline matters because model usage logs can reveal productivity shifts before they show up in official statistics. Aghion's creative destruction framework matters because it predicts which sectors restructure first. Agrawal's work on scientific discovery matters because AI may change the rate of invention itself.

The atlas publishes aggregated, anonymized usage data. Researchers, policymakers, and workers can explore the same dashboards the team uses. The goal is a shared evidence base rather than competing narratives.

How the research connects to policy and training

The announcement frames the expanded team as a scientific bridge. Their output will directly inform future ATLAS updates and empirical publications. The bridge has three concrete targets.

First, identify organizational practices that let AI upskill workers instead of displacing them. Second, map public policy frameworks that spread gains broadly rather than concentrating them. Third, design training programs that democratize expertise so the productivity boost reaches beyond early adopters.

Advising the World Economic Forum and the United Nations gives Madgavkar a channel into global policy discussions. Rock's econometric rigor means the claims will withstand peer review. Aghion's macro models will project long-term trajectories, not just quarterly snapshots. Agrawal's focus on the innovation frontier links today's tooling to tomorrow's discovery rate.

Policy is not set by the program. It supplies the measurement layer that policy arguments currently lack. In the source document's words, the team serves as a scientific bridge directly informing future ATLAS updates and empirical research.

What this changes for anyone using AI at work

If you use Google's AI tools — Gemini in Workspace, NotebookLM, developer APIs — your anonymized usage patterns already feed the atlas. The expansion means more economists will interpret those patterns, and more dimensions will be measured.

The team will track how adoption differs by company size, geography, and industry. They will correlate usage intensity with productivity metrics where data permits. They will watch for skill upgrading versus deskilling within occupations. They will compare diffusion speeds across countries.

None of this produces an individual report card. The atlas shows aggregates. But the aggregates shape the training programs your employer might buy, the policy incentives your government might offer, and the product roadmaps Google itself might prioritize.

The document does not say how many users currently feed the atlas. It does not specify which productivity metrics will be correlated with telemetry. It does not name the small business surveys Madgavkar will run. Those details will appear in future publications at ai.google/economy.

The mechanism is straightforward: better measurement enables better decisions. The expansion adds the measurement capacity Google judged necessary for the next phase.