AI-DRAFTED BILLS CLOG HOUSE LEGISLATIVE COUNSEL

Congressional lawyers are struggling to manage a surge of legislative proposals created with artificial intelligence tools. The influx has created an unexpected bottleneck: reviewing and correcting AI-drafted bills now consumes more staff resources than traditional drafting would require. The Legislative Counsel's office, responsible for ensuring bills meet constitutional and procedural standards, has become a de facto quality control operation for AI-generated text. Sources indicate the corrections needed are substantial enough to negate any efficiency gains from automated drafting. This dynamic reflects a broader challenge as lawmakers experiment with AI tools to accelerate legislative work. While proponents argue AI can speed up the bill-writing process, the reality appears more complex. AI language models can generate plausible-sounding legislative language, but often miss technical requirements, create logical inconsistencies, or violate existing legal frameworks that trained legislative counsel would catch naturally. The backlog raises questions about whether adopting AI for bill drafting actually improves congressional productivity. Rather than reducing workload, the technology appears to be redistributing it—shifting effort from initial composition to extensive revision and repair. No timeline has been provided for when the Legislative Counsel's office might adapt to this new workflow, or whether they plan to implement systems to better screen AI drafts before submission. The situation also highlights a gap between AI capability and specialized domain expertise: while the tools can produce voluminous text quickly, they lack the institutional knowledge required for legislative drafting. The issue comes as Congress grapples with how to govern AI more broadly, even as its own internal operations struggle with the technology's limitations.