OpenAI & Anthropic Respond to Trump's AI Security Order

OpenAI and Anthropic have responded to a new executive order concerning AI security. This order directs federal agencies to secure voluntary agreements for testing AI models with developers. Here's the thing: The government aims to establish a framework where developers can voluntarily provide early access to advanced AI models. This is specifically for models with major cyber capabilities, before they are released to the public. The goal is to boost AI cyber defenses. What's interesting is that the US President also directed several government departments and officials to secure these testing agreements. The order also calls for an AI cybersecurity clearinghouse to coordinate vulnerability scanning and remediation. OpenAI CEO Sam Altman welcomed this approach, stating the US must lead in AI development and safety. Anthropic also welcomed the decision, calling it necessary for strengthening America's leadership in AI. They look forward to collaborating on its implementation. However, critics like Senator Bernie Sanders called for more robust congressional action. He believes the executive order is voluntary and does little to protect Americans. This matters because it highlights ongoing debate about how best to manage the rapid development of artificial intelligence.