Florida Republican claimed to use Claude as a proofreader, nothing else.
If there's one thing we love more than catching a politician doing something silly, it's the excuse they confect to try and get out of it. The latest involves Florida Republican Anna Paulina Luna, who was caught using AI in a draft amendment to a bill because the text included the phrase "Claude responded:." Which might hint that someone pasted in a conversation with the Anthropic chatbot of the same name and forgot to hide it. Luna was quick to shut down the accusation, posting on X (as reported by Gizmodo), that her staff "used AI to correct a draft text and didn't edit," adding that it's "not a shocker" as "most staff use it." Not long after, her posts were deleted, probably because it wouldn't be ideal to admit you used AI to do your job for you. She then posted an amended explanation, saying "my staff used AI to spell/grammar check the amendment SUMMARY, not the actual amendment text itself." Shortly after, she posted again saying "FYI NO Legislation is ever drafted with AI," and that screenshot was merely depicting an "AI summary of the bill that's also used for spellcheck." She also asked "what dork planted this story?" and, to further bolster her reasonable credentials, threw in a tears of joy emoji for good measure.
Now, look, I'm just a simple, country lawyer with none of your highfaultin' fancy ways, so I entirely believe Luna's response. After all, it's very reasonable for someone to use an AI chatbot to check their spelling and grammar, since that feature hasn't been widely available on computers since Luna was six years old. Hell, even if you used Superhuman's (Grammarly) brand of ghoulish AI to check your spelling and grammar, it wouldn't insert references to its brand into the text of the documents you're working on. Which makes me think that, you know, maybe parts of this story don't quite check out.