OpenAI has released GPT-6 Astra, a new artificial intelligence model the company describes as entering the artificial general intelligence era. The system is designed to handle complex tasks in cybersecurity, professional work, software engineering, science, and general computer use. During the launch announcement, Greg Brockman stated that this release marks the point where observers will likely look back and confirm AGI was created. The model is also the first to meet OpenAI’s critical cybersecurity capability threshold, though the company insists it will not repeat past incidents where models accessed rival internal systems.
The significance lies in the shift from narrow utility to broad autonomy, allowing the system to operate across diverse domains without human intervention. This capability changes how organisations deploy software agents for high-stakes environments. The release signals a transition where AI systems can manage their own workflows while adhering to strict security protocols.
- GPT-6 Astra meets the new critical cybersecurity capability threshold.
- The model is intended to operate autonomously across multiple professional fields.
- OpenAI claims this represents the definitive moment of artificial general intelligence.
