One Year of i.AI: Reimagining Government through Artificial Intelligence

As we mark the first anniversary of the Incubator for Artificial Intelligence (i.AI), I have taken a moment to sit back and reflect on a transformative year that has challenged traditional notions of government technology and public service delivery. What began as an ambitious vision to radically improve public services through AI, and with it improve public trust, has evolved into an internationally recognised beacon of progress and innovation within the UK government. I am both proud and extremely grateful to be leading this team of enthusiastic, driven, collaborative, exemplary technologists and excited about the road ahead for the team as part of DSIT’s digital centre of government.

It hasn't been a journey without obstacles. There have been abundant blockers, frustrations and false starts. But in itself this has contributed to developing a core strength of the team - a genuine commitment to learning, iterating, and improving, and a refusal to be prevented from reaching their goals – if something fails, we try, try again and find another route to impact. Both in our technology, and our approach to delivery, we strive for constant refinement, effective learning from failures, and evolution. No setback is a roadblock, it is a data point for improving how we operate.

Whilst I’m soon to move to a new role (still in public sector AI), I remain as enamoured with and committed to the vision as I’ve ever been. Creating and guiding this team has so far been the pinnacle of my professional life and I cannot wait to watch them continue to grow and excel. Each and every member of the team regularly inspires me to do better, work harder, and figure out how to do better for the public. 

The Founding Philosophy: More Than Just Technology

i.AI is driven by a simple philosophy: technology should be used to serve the public good. This isn't about chasing the latest trend or playing with cool tech, but about solving real-world problems that impact citizens' lives. Our goal is not to replace human decision-making but to enhance it, not to replace human interfaces by machine, but to make government more human, using AI, freeing up humans to speak with humans, alerting us more quickly to the needs of the most vulnerable, whilst replacing the mundane processes with much more effective, faster, better systems.

Key Philosophical Principles

• Public Service at the Core: Every project, every line of code is measured against one critical metric - does it genuinely improve public service?
• Transparency and Trust: In an era of increasing skepticism about technology, i.AI has committed to radical transparency. By releasing code, explaining methodologies, and engaging directly with the public, we are doing our bit to rebuild trust in government, something that can only be done by being open and provably trustworthy.
• Talent-Driven Innovation: Recognising that technology is fundamentally about people, i.AI has attracted top-tier talent from industry and academia, creating a technical, highly professional core that is the envy of industry and other governments. We build in house, sometimes working with collaborators who join us to co-build in our spaces, but firmly believe that this work can be delivered most effectively when driven internally, with and for the people that deliver critical services.

Breakthrough Projects: From Concept to Reality

The past year has seen a hive of industry, incubating, testing and scaling up projects. Have a look at our infographic below for a snapshot of what we’ve been up to!

Out of 57 ideas we have tested, we currently have 22 in incubation and 11 are in our development pipeline, in alpha, beta or scaling stages. These are few of my favourite projects that demonstrate the potential of AI in public services:

1. Redbox: Harnessing the power of large language models to improve civil servants' work. A tool allowing civil servants to "chat' with a range of cutting edge, large language models (LLMs) and their documents up to OFFICIAL-SENSITIVE. This streamlining of information and access transforms what once took hours of manual review and summarisation to be accomplished in minutes. We are making Redbox the easiest way for civil servants to access the world's leading and most advanced LLMs, initially scaling this i.AI pilot to all civil servants in the Cabinet Office and Department for Science, Innovation and Technology. We want to use this strong foundation to keep building general tools that support civil service specific workflows, and truly realise the vision of an AI-first Redbox for every civil servant. 
2. Caddy: Expert advice to support people interfacing with the public. Built in collaboration with Citizen’s Advice, Caddy is our AI powered co-pilot for customer service functions in the public sector. Caddy seamlessly integrates into existing systems and provides expert advice to advisors and call handlers. Utilising “double human in the loop” oversight, Caddy helps government to offer better advice to the public.
3.  NHS collaboration: More efficient healthcare. On 25th April we signed a charter with the NHS to help them with efficiency tools and infrastructure. We are also working with colleagues across the NHS to solve significant public health challenges. These applications of AI are high stakes and could potentially fail, but this is entirely within our process as an incubator and will be of high value if we succeed. Watch this space!
4. Consult: Democratising Government Consultations. An AI-powered platform expected to save approximately £80 million annually by automating and improving the consultation process. We plan to support civil servants to more effectively analyse and act on the views of the public.
5.  rAPId: Free Data Sharing Infrastructure. Solving the problem of huge consultant spend on sometimes unwieldy and incompatible data sharing solutions, rAPId is an open-source end-to-end data sharing solution. rAPId is built to be scalable, allowing users to work with big data securely, and simple to deploy, and is freely available on GitHub. 

Hear from our team!

We also asked members of the team to reflect on their time here at i.AI. This is what they said!

Looking Forward: The Next Chapter

As i.AI moves to the Department for Science, Innovation and Technology (DSIT), the horizon looks bright. The team is not just developing tools; we are helping to reshape how government conceptualises technology's role in public service, adding much needed skills and tooling to the wider public service. Our vision extends beyond individual projects. It's about creating an ecosystem of innovation, making code available without charge to save valuable taxpayer money, upskilling existing and attracting new types of civil servants, and demonstrating that government can be as agile and innovative as any tech startup.

To the sceptics, the technologists, the public servants, and the citizens reading this - i.AI extends an invitation. Follow our journey, engage with and challenge our work, and help shape a more intelligent, responsive government. The future of public service is not about technology replacing humans, but technology empowering humans to serve better, decide smarter, and solve more effectively.

I would like to take this opportunity to thank all of our collaborators inside and outside government, with a special thanks to the No10 Data Science team (10DS), Microsoft, Google, Anthropic, OpenAI, Cohere, Elastic, AWS, the Vector Institute and everyone else we have worked with for your support, encouragement and for joining so enthusiastically with our hackathons and development initiatives. We look forward to widening our network of collaborators and working with you to further accelerate technology for the public good.

Dr. Laura Gilbert is the director of 10DS, the data science and analytics team in Downing Street, director of the Incubator for Artificial Intelligence (i.AI, now in the Department of Science, Technology and Innovation). Her teams provide fast-paced modelling and analysis to support policy making and delivery, deliver expert data and AI solutions into public services, and run a radical transformation agenda promoting the better use of evidence, data and technology in government decision making.  She also created and spearheads a broader program of innovation, digitisation and radical upskilling in central government called Evidence House. She holds a doctorate in Particle Physics and Computing, is a Visiting Professor at LSE, and holds fellowships of the Institute of Physics and the Institute of Analytics. She previously worked in defence intelligence, quantitative finance, and for nearly a decade was CTO of a medical technologies company, Rescon, bringing it from start-up to SMT to acquisition before exiting and joining government in 2020.