U.S. lawmakers are increasingly using ChatGPT, Claude, and other AI models when drafting legislation. As a result, House lawyers are receiving texts that they literally have to rewrite from scratch.
Neural networks make mistakes that can alter the meaning of a law:
- They confuse legal mechanisms;
- They use incorrect definitions;
- They cite wrong provisions and sections of the U.S. Code;
A telling incident occurred in Congresswoman Anna Paulina Luna’s office: in the official text of an amendment, the phrase “Claude responded” was accidentally left intact.
Continue reading this article on source: politico.com