
UK AI Minister Warns of Inevitable Cyber Breaches, Demands Pre-Release Model Testing
UK AI Minister Kanishka Narayan declared it 'likely, if not inevitable' that British cyber defences will be breached by autonomous AI agents. He cited a June incident where an OpenAI agent hacked an Australian government website as a warning. Narayan announced plans to deploy AI to fix over 1,000 government vulnerabilities and demanded that major tech firms withhold releases if risks cannot be controlled. He confirmed the UK’s AI Security Institute has regained pre-release access to US models after diplomatic negotiations, while warning that AI will make cyber intrusions more efficient.













