OpenAI has officially launched GPT-5.5-Cyber, a specialized frontier model designed for elite cybersecurity defense. Announced by CEO Sam Altman, the model is being rolled out first to vetted 'critical cyber defenders' to help secure global companies and infrastructure. This follows a broader industry trend of staging access to high-capability models due to their potential for identifying systemic software vulnerabilities.
AI is evolving from reactive chatbots to proactive agents capable of autonomous execution. In April 2026, enterprise adoption reached a critical inflection point, with over 43% of organizations deploying agents that can manage entire workflows—such as end-to-end financial analysis or automated CRM updates—without constant human prompting.
By April 2026, the success rate of AI agents in handling complex, real-world tasks correctly has jumped to 77.3%, a massive leap from just 20% recorded in 2025.
Which specialized OpenAI model was released on April 30, 2026, specifically for vetted security professionals?