Fed’s Waller says FRED now holds over 850,000 data series, with more AI tools planned

The Federal Reserve’s most-used free database keeps getting bigger. Speaking at FRED Con 2026, Governor Christopher J. Waller said the Federal Reserve Economic Data platform, better known as FRED, now hosts over 850,000 data series.

He also said the platform will integrate more AI tools to improve how people find and rely on that data.

What Waller laid out at FRED Con

Waller delivered the opening remarks at FRED Con, a one-day conference hosted by the Federal Reserve Bank of St. Louis. The event is scheduled for October 1, 2026, running from 9 am to 4 pm CT.

It is livestreamed on the St. Louis Fed’s YouTube channel.

The headline figure was scale. FRED’s library of more than 850,000 series pulls from a range of sources, including the Bureau of Economic Analysis, the Bureau of Labor Statistics and the Census Bureau.

The second part of Waller’s message concerned artificial intelligence. Plans call for expanding AI tools on the platform, with the stated goal of improving both data accessibility and reliability.

A conference built around trust, AI and storytelling

The FRED Con agenda tracks closely with the themes Waller raised. Panels cover generative AI’s role in economic data discovery and analysis, along with trusted data practices that emphasize accuracy and transparency.

Other sessions focus on storytelling with numbers and on accessibility for diverse users.

The speaker lineup reaches well beyond the Fed. Representatives from Amazon, FedEx, The Economist and the US Census Bureau are taking part, alongside staff from the St. Louis Fed.

Waller’s longer-running interest in AI

This is not the first time Waller has talked about artificial intelligence. He has previously addressed AI in the context of payments innovation and the Federal Reserve’s own internal operations.

His FRED Con remarks extend that thread from how the Fed works internally to how it serves the public.

What this means for FRED users

For the people who use FRED daily, the most immediate implication is that discovery could get easier. With over 850,000 series, finding the exact dataset you need can feel like searching for one specific receipt in a shoebox the size of a warehouse.

The flip side is risk. Generative AI tools are known for producing confident answers that are sometimes wrong, and a mistake attached to a Federal Reserve brand carries extra weight.

That is presumably why the conference pairs its AI panels with a dedicated focus on trusted data practices, accuracy and transparency.