London sets out ‘city innovation stack’ for AI and digital public services

London is to develop a new city-wide approach to digital infrastructure and AI as it looks to move public sector innovation beyond individual pilots and reduce duplication between organisations.

Theo Blackwell, London’s chief digital officer, has set out a four-layer “city innovation stack” covering digital infrastructure, collaboration, AI deployment and public trust.

The approach is intended to help London’s public bodies share data, technology and standards rather than repeatedly developing and procuring similar capabilities independently. It forms part of the London Growth Plan and builds on the Mayor’s previous Smart London programmes.

Blackwell said London’s combination of 32 local authorities, NHS organisations, Transport for London (TfL), emergency services and other public bodies gives it considerable potential to develop and scale technology collaboratively. The challenge, he argued, is turning that scale into practical benefits for residents.

The proposed stack starts with a foundation of data, connectivity and computing infrastructure. London plans to expand its Data for London Library, improve gigabit and mobile connectivity and align the growth of data centres with the capital’s decarbonisation plans.

The Data for London Library now brings together 20,000 urban datasets, while gigabit coverage has increased from four percent of homes and businesses in 2017 to 92 percent today, according to Blackwell. London also has 99 datacentre sites, with demand expected to grow further as adoption of AI increases.

Common technology across public services

The second part of the strategy focuses on collaboration between public bodies.

Blackwell said London boroughs currently spend between £7 million and £20 million each year on areas of commodity technology such as networking, hosting, security and devices, despite having many of the same requirements.

Working with GDS Local, London wants to explore common approaches to these technologies as well as reusable components for functions including payments, appointment booking, forms, identity verification and case management.

The approach reflects wider government efforts to encourage public bodies to reuse common technology rather than independently buying or building similar capabilities.

London also plans to expand the role of the London Office of Technology and Innovation (LOTI), which has delivered more than 60 collaborative projects since 2019 and established an AI network involving 200 officers.

Under what Blackwell described as “LOTI 3.0”, the organisation would gain dedicated research and development capacity to test new approaches to local public services at greater scale.

The model could also extend into health. Through the NHS London-funded Health Innovation Zone, the aim is to make it easier for technologies successfully tested by one NHS trust to be adopted by others without repeating assurance processes.

GLA to establish central AI and technology function

AI forms the third layer of the strategy, covering predictive analytics, generative AI, agentic AI and autonomous technologies including drones and self-driving vehicles.

Generative AI is already being used across London public services for areas including housing complaints, consultation analysis and casework, while applications within the NHS include diagnostics, triage and administration.

A significant organisational change will see the Mayor establish a central AI, digital, data and technology function across the GLA Group. The function will set direction for technology investment, identify opportunities and coordinate adoption across the group, rather than replacing existing delivery capabilities.

Blackwell said the intention is to avoid organisations separately addressing many of the same questions around AI ethics, skills and procurement, instead allowing them to share standards and expertise.

London is also preparing for greater use of agentic AI, as well as autonomous vehicles and drones. The strategy proposes shared governance for autonomous systems as their use becomes more widespread.

Drones are already used by TfL for infrastructure inspection, London Fire Brigade to assess fires and the Metropolitan Police for incident response. Commercial self-driving vehicle pilots are also planned for 2026/27.

Building public trust

The final layer of the stack focuses on trust, including public participation in decisions about technology, AI literacy and the impact of automation on employment.

London plans to develop an AI literacy offer to help residents understand the capabilities and limitations of AI-enabled systems, building on the existing Get Online London digital inclusion programme.

Blackwell also highlighted the challenge of AI-generated disinformation and said London wants to make authoritative public information easier for AI systems to find and interpret.

That could include structuring key datasets covering areas such as crime, health, housing and economic performance, alongside content published on london.gov.uk, so that AI-generated answers about the capital are more likely to draw on authoritative sources.

Blackwell said the goal is to deliver more joined-up services for Londoners while giving public servants shared infrastructure and tools rather than requiring individual organisations to repeatedly build or procure the same capabilities.