According to the OpenAI announcement, the company released a major update to Model Spec, which defines the standards of behavior for AI models. The new version emphasizes customizability, transparency, and freedom of thought, ensuring that users are free to explore, discuss, and create, while maintaining the necessary security safeguards. Updates include: • Model instruction priority: clarify the instruction priority of the platform, developers and users, and improve customization capabilities. • Information authenticity: The model avoids biased guidance, encourages objective discussion, and helps users make independent decisions. • Security perimeter: Set standards for refusing to execute high-risk requests, such as harmful behavior or privacy violations. • Knowledge Sharing: The specification has been open-sourced to the public domain, allowing developers to use, modify, and build freely. OpenAI says the specification will continue to evolve and welcomes community feedback.