OpenAI's New 'Astra' AI Capable of Autonomous Cyberattack Development

OpenAI has announced that its new AI model, Astra, has reached the company's 'Critical' cybersecurity threshold. This designation indicates that the model is capable of identifying previously unknown vulnerabilities and autonomously developing methods to exploit them within hardened systems without human intervention.