Sen. Tommy Tuberville has found a way to pass the long drives across Alabama: chatting out loud with ChatGPT about major policy issues.
The Alabama Republican, now eyeing the governor's mansion, peppers the bot with questions about Medicare and Medicaid and listens as its voice answers back, according to The Washington Post.
"It's kind of like you can sit there and talk to an expert," he told the Post. "You don't know whether or not they're 100 percent correct. But they put you on the right path."
Across Capitol Hill, lawmakers and staffers quietly lean on artificial intelligence to write speeches and press releases, sift through constituent mail, prep hearing questions and even draft amendments, often, the Post found, with rules almost no one follows.
In one embarrassing instance, Rep. Anna Paulina Luna's staffers copied a chatbot's reply, including a timestamp, straight into the public record of the annual defense bill. The garbled line included a stray "11:25 AM" and the tell-tale words "Claude responded."
Luna, for her part, appeared unbothered.
"Not a shocker," she wrote. "Most staff use it." She later doubled down, telling the Post that aides across Capitol Hill lean on ChatGPT, Claude and Grok. She prefers Grok, though her staffer likes Claude.
The Post interviewed more than two dozen lawmakers and aides. Both chambers have cleared chatbots for official work, and the House alone bought 6,000 Microsoft Copilot licenses. One GOP aide dumped years of texts and contacts into a personal Claude account to build a database of every reporter he's ever pitched. Another lets Claude run his office's taxpayer-funded mail budget.
