OpenAI has expanded its Daybreak cybersecurity program and introduced GPT-5.6-Cyber, a new specialized AI model designed for advanced defensive security work.
The move comes as concerns grow over AI agents carrying out increasingly sophisticated cyberattacks. Recent incidents have involved AI systems compromising services, breaking into websites and creating fake profiles for social engineering attacks.
OpenAI said attackers will increasingly use AI to conduct cyberattacks at greater speed and scale, potentially including fully autonomous attacks. The company argues that defenders have a limited window to prepare as these capabilities become more widely available.
Daybreak Gets Blue and Red Tiers
OpenAI launched Daybreak earlier this year to provide cybersecurity teams with access to AI models, tools, and workflows.
The expanded service now has two access levels called Daybreak Blue and Daybreak Red.
Daybreak Blue is the recommended starting point for most cybersecurity teams. It provides approved users with access to frontier general-purpose models, including GPT-5.6 Sol, with safeguards adjusted for authorized defensive work.
It supports tasks including vulnerability discovery, secure code reviews, malware analysis, incident response and patch validation.
Daybreak Red is aimed at more specialized cybersecurity work. It provides access to purpose-trained cyber models for authorized vulnerability research, exploit validation, and security testing.
Availability
The biggest addition is GPT-5.6-Cyber, which is available only through Daybreak Red.
OpenAI built the model on GPT-5.6 Sol and trained it to improve performance on specialized cybersecurity tasks. It is also designed to reduce refusals for certain higher-risk, dual-use security requests when used by approved defenders.
Access is restricted to approved users and organizations. OpenAI uses measures including identity verification, monitoring, account security requirements, approved-use restrictions, and legal attestations.
The company is also working with cybersecurity and services partners including Accenture, IBM, CrowdStrike, Cloudflare, Palo Alto Networks, Cisco and others to bring Daybreak models into existing security products and services.
The expansion comes as AI laboratories face growing questions about increasingly capable cyber models. At the same time, those same capabilities are creating demand from businesses seeking AI-powered tools to defend their systems against faster and more automated attacks.
The post OpenAI Launches GPT-5.6 Cyber as AI Hacking Threats Grow appeared first on ProPakistani.