Lawmakers in the United States are increasingly turning to basic versions of ChatGPT and Claude to draft legislation, with a troubling result: AI-generated bills containing writing errors have surged 70%, according to Tecnoblog.
The problem extends beyond simple typos. These documents suffer from incorrect information and lack the depth needed for sound policymaking. The consequences are tangible—frequent revisions have become necessary to catch errors, slowing down legislative processes and adding unnecessary delays to bill review and approval cycles.
The trend highlights a broader tension in government: the appeal of speed and cost savings from AI tools versus the complexity and precision required for legal documents. While AI writing assistants can draft text quickly, they're clearly not yet ready to replace skilled legislative writers, especially when accuracy and legal clarity are paramount.