AI Bots Now Make Up Half of Traffic to US Fed’s FRED Database
Artificial-intelligence agents are rapidly becoming major consumers of US economic data, forcing the Federal Reserve to redesign one of its most widely used public services.
Topics
News
- Fired OpenAI Researchers Warn Dismissals Could Chill Safety Work
- ARTEX Developer Closes AI Hacking Tool After Bank Attacks
- AI Redraws Global Race for Quant Talent as Coding Tools Lower Entry Barriers
- AI Dispatch | India Draws Billions Into AI
- AI Bots Now Make Up Half of Traffic to US Fed’s FRED Database
- Musk Blames ‘Oligarchs’ For Delay in Starlink’s India launch
Artificial-intelligence systems now account for about half of visits to the Federal Reserve’s FRED economic database, a surge that is forcing the US central bank to rethink how one of the world’s most widely used sources of economic data serves machines as well as humans.
Traffic to FRED is growing about 150% a year, with most of the increase coming from AI systems and automated bots, Federal Reserve governor Christopher Waller said at a conference hosted by the St. Louis Fed last Thursday, 1 October.
“Already about half of visits to FRED are from AI agents retrieving data, with the other half from flesh-and-blood visitors,” Waller said.
FRED, short for Federal Reserve Economic Data, hosts hundreds of thousands of data series covering inflation, employment, interest rates, economic output and financial markets.
The site was designed primarily for researchers, investors, journalists and other human users. AI assistants are changing that bargain by retrieving data automatically and, increasingly, presenting it to users without their ever visiting the underlying website.
That creates problems beyond server capacity.
Waller said FRED’s managers cannot always determine whether AI agents correctly attribute figures to the original provider or to FRED itself. Automated systems can also misinterpret statistics, confuse relationships between variables or present revised data without making clear what was known at an earlier point in time.
The St. Louis Fed has responded by launching an official FRED connector based on the Model Context Protocol, or MCP, which allows AI applications to query the database in a more structured way.
The approach is notable because FRED has historically restricted disruptive scraping. Its terms prohibit automated extraction that harms the stability or performance of the service and also impose restrictions on using FRED content for training machine-learning systems.
The traffic surge comes as financial companies are simultaneously wrestling with the security risks created by more capable AI agents.
JPMorgan Chase Chief Executive Jamie Dimon said this month that cyber risks had increased “10-fold” after Anthropic developed its advanced Mythos model, arguing that AI had created vulnerabilities companies previously did not know existed.
