New experts join Google’s AI & Economy team
Sep 18, 2026, 7:00 AM · Google AI Blog
Google expands its AI & Economy program with Nobel laureate Philippe Aghion, Ajay Agrawal, and directors Anu Madgavkar and Daniel Rock—pairing ATLAS adoption data with heavier academic firepower on work, growth, and diffusion.
Why it matters
Tracking who uses Gemini at work isn’t enough once AI is reshaping jobs and firms. Google is staffing the AI & Economy Research Program like a serious macro and labor shop, not a blog team.
Aghion (2025 Nobel) joins Michael Spence and Diane Coyle as advisors; Agrawal joins David Autor’s visiting-fellow lane. Madgavkar (ex-McKinsey Global Institute) and Wharton’s Daniel Rock will direct empirical work with DeepMind’s Alex Imas and Google’s Zanna Iscenko.
That bench can inform ATLAS updates and policy talk—or become a soft power channel for Google’s preferred framing of shared prosperity. Both can be true.
From the desk
We’re glad the economic measurement is getting real economists—and we’re watching who the research serves.
ATLAS v1.0 already maps how people use Google’s AI tools. The expansion aims at future of work, productivity, global diffusion, and scientific discovery. Aghion’s innovation-and-creative-destruction lens on AI’s long-run growth path, Agrawal on AI economics and robotics with Autor, Madgavkar on labor markets and small-business ecosystems, Rock bridging model telemetry to econometrics—that’s a serious agenda.
Useful AI needs exactly this kind of evidence: which practices upskill workers, which policies spread gains, where disruption concentrates. The downside if the program stays mostly a Google-facing narrative shop: policymakers get polished charts without adversarial scrutiny of concentration, labor bargaining, or energy costs.
I’m watching whether future ATLAS releases and peer-facing papers change public debate—or mainly decorate Google’s AI-economy brand.
Context
Google AI Blog post dated September 18, 2026, by Scott Strand and Zanna Iscenko. Program sits across Technology & Society and the Chief Economist’s Office, with DeepMind’s AGI Economics director in the leadership mix.
Who feels it
- Policymakers and multilaterals
- A better-staffed Google research channel on AI labor and diffusion will shape briefings even when independent academia disagrees.
- Workers and small businesses
- Empirical work on generative AI’s workforce and SMB effects could surface practices worth copying—or harms worth regulating.
- Academic economists
- Access to Google telemetry via fellows and directors may accelerate papers—and raise questions about independence.
What to watch
- Next ATLAS update informed by the expanded team.
- Published empirical papers from Madgavkar and Rock’s streams.
- Whether findings push Google product or public-policy positions in measurable ways.
Companies: Google