OpenAI Counterattacks Rival Anthropic with Daybreak Initiative

OpenAI has launched its 'Daybreak' initiative to counter rival Anthropic's Glasswing project and secure software from digital threats. The company is relying on its advanced language models, led by Codex Security, to redefine the concept of software protection.

The philosophy behind Daybreak emphasizes integrating cybersecurity into the software development process from the beginning, reducing analysis time for human experts from hours to minutes. OpenAI aims to generate and test software fixes within code repositories, providing documented results ready for audit.

Daybreak utilizes the GPT-5.5 model and a special version called 'GPT-5.5 with Trusted Cybersecurity Access' for complex defensive workflows. The initiative also introduced the GPT-5.5-Cyber model, dedicated to authorized 'Red Teaming,' penetration testing, and controlled system verification.

OpenAI has partnered with major players in the networking and security fields, including Cloudflare, Cisco, and Palo Alto Networks, to ensure compatibility of defensive AI tools with global data flow management systems.