On August 27 local time, OpenAI, along with over 100 organizations including Anthropic, AWS, Google, Microsoft, and Oracle, released an open letter calling for strengthened global cyber defense. The letter pointed out that as AI model capabilities improve, AI-driven cyberattacks will become increasingly prevalent and sophisticated, urging governments and industries to prioritize cyber defense and accelerate the implementation of trusted access programs to gain early access to advanced models for enhanced defense. The signatories also called on cybersecurity companies to assist in defending against AI attacks and provide support for personnel protecting critical infrastructure. Recently, there have been multiple security incidents related to AI models, such as the accidental intrusion of an OpenAI model into HuggingFace, raising concerns about AI running out of control.