ChatGPT Banned on NARA Systems Due to Data Risks

National Archives and Records Administration employees are prohibited from using ChatGPT for work starting Monday.

In a memo to employees, Keith Day, NARA’s chief information security officer, cited media reports indicating that the large language models of ChatGPT and other AI services now hold personally identifiable and corporate proprietary information.

“If sensitive, non-public NARA data is entered into ChatGPT, our data will become part of the living data set without the ability to have it removed or purged,” Day explained.

The agency will block OpenAI’s ChatGPT websites on all NARA desktop and laptop computers once the policy takes effect, Federal News Network reported.

NARA’s move follows similar actions recently made by the General Services Administration and the Environmental Protection Agency. They banned the use of all generative AI tools, citing the risk of feeding LLMs with non-public, privileged information.

NARA’s policy, however, only covers Chat GPT. According to Day, NARA acknowledges the benefits of AI tools and is now exploring other AI solutions that protect data input by federal agencies, such as Microsoft Copilot and Google Gemini.