Claude & ChatGPT Draft Laws on Capitol Hill: 8 Officials Cite Error-Filled Bills
General-purpose models from Anthropic and OpenAI are now feeding raw output into the U.S. legislative process, and the House Office of Legislative Counsel is struggling to keep up. Eight current and former officials say Claude and ChatGPT drafts arrive riddled with hallucinated statutes and incorrect legal definitions, forcing lawyers into heavy rewrites. The episode is a real-world stress test of where LLMs fail in high-precision, verifiable text generation.
Source: townhall.com · wcbm.com