Florida Republican Anna Paulina Luna sparked controversy after a draft amendment displayed the line “Claude responded:” suggesting a copy‑paste from Anthropic’s chatbot Claude. Luna responded on X, saying her staff “used AI to correct a draft text and didn’t edit,” and added that most congressional aides rely on such tools. The incident quickly drew media attention.

Within minutes Luna deleted the original posts, then posted an amended explanation claiming the AI was applied only to the amendment’s summary for spelling and grammar, not the bill’s substantive language. She later insisted “NO Legislation is ever drafted with AI,” framing the screenshot as merely an AI‑generated summary used for proofing. The back‑and‑forth highlighted how quickly officials manage digital missteps.

The episode underscores that congressional staff increasingly treat AI assistants as ordinary spell‑check utilities, a practice that blurs the line between routine editing and substantive policy drafting. As lawmakers adopt tools like Claude, transparency about where and how AI influences official text becomes a matter of public accountability, not just a technical footnote.

Critics argue that even simple proofreading can embed AI branding into public records, potentially confusing constituents about the origin of legislative language. Whether the practice prompts new disclosure rules remains to be seen, but the incident already fuels debate over AI’s role in the lawmaking process.