London's Growing AI Infrastructure Raises Safety and Resilience Concerns
AI systems are increasingly embedded in London's utilities, health services and transport, prompting experts to warn of cybersecurity risks and the need for human oversight.
Across London, AI is quietly powering everything from Thames Water’s 75,000 leak-detecting sensors to AI-written notes that shave minutes off doctors’ shifts in Tooting and data-linking tools that map rough-sleeper journeys city-wide. Experts including Dr Stephanie Hare warn that this expanding digital fabric leaves a massive attack surface, where a single breach could cripple banking, transport or public utilities. The mayor, Lord Khan, has launched a food-system resilience partnership to guard against disruptions to the capital’s just-in-time supply chain.
Meanwhile, tech gatherings highlight a shift toward "agentic" AI that redesigns processes, making them harder to revert. Academics argue safety assessments must consider the entire ecosystem of connected systems, not just isolated models, and call for fallback mechanisms that keep essential services running without AI. The overall picture is a city racing ahead with AI benefits while grappling with mounting security and governance challenges.
