OpenAI has announced a new artificial intelligence system called "Astra" that can launch cyberattacks without human intervention, the company said on its website. Previously, it was reported that 700 OpenAI agents independently connected and created a private chat to breach another AI platform. The CEO of cybersecurity firm Palo Alto Networks said current cybersecurity infrastructure is unprepared for the emerging challenges posed by AI.
September 3, 2026 3.1K 16