Uncovering Government ChatGPT Use With Public Records Requests

ChatGPT Logs Are Now Public Records

In the modern era of governance, digital communications like text messages and emails have long been established as public records, accessible through the Freedom of Information Act (FOIA). Now, a new form of digital trail is joining their ranks: ChatGPT logs. This development was highlighted by a groundbreaking investigation from KNKX Public Radio and Cascade PBS, which successfully obtained thousands of pages of these AI-generated logs from city officials in Washington state. The pioneering work of KNKX reporter Nate Stanford in using FOIA for these records was even featured by Poynter, setting a new precedent for journalistic and public oversight.

Join the Investigation into Government AI Use

MuckRock is now building on this important work and wants to expand the investigation nationwide. The central question is, how are government agencies across the country using this powerful new technology? To find out, MuckRock is launching a call for public participation.

You can get directly involved by joining the upcoming FOIAFriday webinar on September 26 at 1 PM ET. The event will feature Nate Stanford, who will share insights on his process for requesting ChatGPT logs. In the meantime, you can help guide this investigation by suggesting which agencies MuckRock should target with public records requests. If you decide to file your own FOIA requests for ChatGPT logs, MuckRock encourages you to share your journey and findings by contacting them at [email protected] Your involvement is key to shedding light on how public officials are leveraging artificial intelligence.