Image: Council on Foreign RelationsThe U.S. Is Losing the AI Credibility War—to Itself
• The U.S. government has abruptly reversed its policy of providing limited access to advanced AI models with cybersecurity capabilities to specific partners. • This shift was triggered by reports that unauthorized parties accessed restricted models and that guardrails in public versions were bypassed in certain instances. • The move is criticized for undermining critical cybersecurity outcomes and damaging the trust between the U.S. government and private frontier AI labs.
cfr.org


