Congressional Staff Misuse AI, Causing Chaos in Bill Text

On August 17, the use of artificial intelligence tools by U.S. congressional staff drew public attention. Staff members from the office of Republican Representative Anna Paulina Luna of Florida directly copied and pasted content generated by Claude, an AI chatbot developed by Anthropic, into the public record of the National Defense Authorization Act, resulting in garbled text appearing in the document. Currently, congressional staff are using AI tools such as OpenAI’s ChatGPT, Anthropic’s Claude, and xAI’s Grok for routine tasks. Existing guidelines governing AI usage are fragmented and lenient, and no staff member has yet faced formal disciplinary action for violating rules related to AI use. Some congressional aides have independently used AI tools without oversight. Sydney Blum, an aide to Democratic Representative Cleo Fields of Louisiana, stated that she personally decided never to use AI to draft documents from scratch, though such a rule is not officially mandated by Congress.