TL;DR

London’s chief digital officer, Theo Blackwell, has set out a four-part plan to stop the capital’s public bodies building and buying the same technology over and over. Its layers run from infrastructure and joint working to AI rollout and public trust, and it includes a new central AI function for the Greater London Authority group. Its most unusual idea is to publish official data in a form AI assistants can find, so their answers about London draw on authoritative sources.

Why London wants a shared approach

The capital has 32 local authorities alongside NHS bodies, Transport for London and the emergency services. Blackwell argues that scale is an asset only if it is used together. By his count, boroughs’ annual spend on commodity technology such as devices, hosting, networks and security runs from £7m to £20m, even though their needs largely overlap.

The plan, which sits within the London Growth Plan, works in four layers:

• Infrastructure: growing the Data for London Library, which now holds 20,000 datasets, improving connectivity and matching data centre growth to the capital’s decarbonisation plans. Blackwell says gigabit coverage has risen from 4% of premises in 2017 to 92%.
• Collaboration: working with GDS Local on shared components for payments, bookings, forms, identity checks and case management, and expanding LOTI, the capital’s office for technology and innovation, into a “LOTI 3.0” with its own research and development capacity.
• AI deployment: a single GLA-wide unit for AI, data, digital and technology, steering investment and adoption, plus shared governance for drones and autonomous vehicles.
• Trust: an AI literacy offer for residents and more public involvement in technology decisions.

Already in use

In London’s public services, generative AI is already used for casework, for analysing consultations and for housing complaints, and NHS uses include diagnostics and triage. Drones are used by TfL for inspections, by London Fire Brigade at fires and by the Metropolitan Police at incidents. A Health Innovation Zone funded by NHS London aims to let one trust adopt technology another has already tested, without repeating the assurance process.

Looking forward

The plan tackles a problem other UK public bodies share. The GDS founder recently warned that public bodies are giving away control over AI, and a shared stack is one way to build that capability centrally. The data idea deserves attention beyond London. Structuring housing, crime, economic and health data so AI systems cite it is a practical response to AI-generated disinformation, though opening data up to machines also calls for care over which systems can reach it and how.