# AI Fed Simulation Favored Holding Rates in Five of Seven October Runs

By Riza Dagoc

Canonical URL: https://www.tokenpost.com/news/regulation/42732
Published: 2026-10-28T16:14:20.000Z
Updated: 2026-10-28T16:14:20.000Z
Section: Regulation

> Sim Fed’s other two runs favored a rate increase. The project’s initial working paper was published in September 2025.

Sim Fed, an experimental AI system created by researchers at George Washington University and Stanford University, favored keeping rates unchanged in five of seven simulations of the Federal Reserve’s October meeting, broadly matching market expectations.

The system’s other two runs favored a rate increase. The simulations were research experiments, not a Federal Reserve decision.

Sim Fed was developed by Tara Sinclair, chair of George Washington University’s economics department, and Sophia Kazinnik, a research scientist at Stanford’s Digital Economy Lab. Sinclair and Kazinnik have spent about 18 months developing the system. They published an initial working paper in September 2025, and the project remains in development.

The Federal Reserve was scheduled to announce its decision at 2 p.m. ET Wednesday (18:00 UTC).