Humans can direct GPT-6 Astra to create a form of controlled Artificial General Intelligence, and it can be made autonomous with appropriate guardrails

Artificial general intelligence: AI matching or surpassing humans across cognitive tasks.

OpenAI's frontier reasoning model, released in September 2026 with a one-million-token context window and state-of-the-art results on computer use, browsing, software engineering and cybersecurity. It is the first model OpenAI rates Critical for cyber capability under its Preparedness Framework, and ships with safeguards that refuse advanced offensive security tasks.