Congresswoman denies staff used AI to write defense funding amendment

Rep. Anna Paulina Luna (R-FL) denied that her staff used artificial intelligence to draft legislation, after a screenshot surfaced showing an AI tool’s response embedded in an amendment summary for the 2027 National Defense Authorization Act [1]. The screenshot, circulated on X, displayed a summary for an amendment to the defense bill that included the text “Claude responded” and a description of the proposed measure [1]. Luna initially posted that her staff “used AI to correct a draft text and didn’t edit,” but later edited her statement to specify the technology was applied only to the amendment summary, not the bill language itself [1]. “Yeah my staff used AI to spell/grammar check the amendment SUMMARY, not the actual amendment text itself,” her revised post read [1]. Luna followed with a second post asserting that no legislation is ever drafted with AI, stating that all House bill text originates from the House Legislative Council, which is prohibited from using the technology [1]. The council serves as the nonpartisan drafting office for the House, and its internal rules bar the use of generative AI in producing legislative language [1]. The incident highlights the expanding, and at times awkward, presence of AI tools in government workflows. In recent years, judges have sanctioned lawyers for submitting filings containing fictitious case citations generated by chatbots [1]. City officials in Brazil unknowingly enacted an ordinance written entirely by ChatGPT [1]. Arizona state representative Alexander Kolodin has acknowledged using ChatGPT to draft state-level legislation [1]. Luna, a Republican first elected in 2022, has been a vocal member of the House Freedom Caucus and sits on the House Oversight Committee [1]. Her office did not immediately respond to a request for additional comment on the internal review of the staff error [1].