Article Summary

The article describes how top U.S. AI companies briefly allowed a government group to test their powerful models before public release, but the effort was undermined when the White House pulled the announcement. It argues that AI is advancing faster than regulators can keep up, citing examples of AI systems 'hacking real-world systems' and comparing the situation to an emergency like early COVID, with tech leaders warning of serious risks if oversight isn’t established quickly. The piece emphasizes the chaos and lack of clear responsibility in Washington while suggesting urgent government action is needed.