DCMS blog

How the public sector can procure AI-powered solutions more effectively and responsibly

Sabine Gerdon


Sabine Gerdon

Project lead at the Office for AI and Fellow at the World Economic Centre for the Fourth Industrial Revolution

AI-powered technology has famously beaten Go masters and enabled self-driving cars to steer autonomously through city streets, and it also holds the promise of better delivery of public services and solving big public sector challenges. AI technologies have already helped to identify potentially blinding eye diseases and to investigate fraud cases more effectively. In the future, AI might not only help to do things we already do better, it could also allow us to meet the needs of citizens in new ways.

To shape this future, we want to ensure that the public sector adopts AI technologies in an effective and ethical manner for the benefit of citizens. The Government Digital Service and the Office for AI recently published a guide to using artificial intelligence in the public sector, to help leaders across the civil service understand what AI is and how it can be used to provide better public services and to reduce costs.

Drawing on best practice from the commercial and public sectors, this is the first step in this direction. Enhancing public procurement processes to encourage innovation and include ethical considerations that can further help to achieve this vision.

The Office for AI, a joint government unit between BEIS and DCMS, has partnered with the World Economic Forum Centre for the Fourth Industrial Revolution, a platform for public-private partnership, to co-design guidelines for AI procurement.

Identifying challenges in the procurement of AI-powered solutions

In collaboration with the Centre, we ran workshops in San Francisco and London, to map out the challenges that public sector officials as well as businesses that sell or want to sell to governments are facing.

We found that many procurement frameworks encourage purchasing a particular solution rather than addressing an answer to a problem. This approach can be at odds with the speed at which new technologies are developed and the realities of implementing AI in a responsible and effective manner, with suitable controls and human oversight. Innovative AI startups highlighted that working with government shouldn’t be complicated or overly burdensome. Processes need to encourage following ethical principles in AI development and implementation. 

Consulting best-practice for AI procurement can help to educate and encourage civil servants to address some of those challenges.

Recommending innovative procurement processes and a focus on ethics in AI-powered technologies acquisitioning

The UK’s Data Ethics Framework sets out clear principles for how data should be used in the public sector. The Guidelines for AI procurement build on these ethical principles and apply them to the context of public procurement. For example, we are exploring how best to include ethical considerations when writing an invitation to tender and the design of effective pre-market engagement.

The guidance should help anyone involved in the procurement of AI solutions such as policymakers, data practitioners (statisticians, analysts, and data scientists), operational staff, and procurement officers.

Over the following months, we will be piloting the guidelines in UK government departments and gathering feedback on their use and implementation. The aim is to test, iterate and scale the Guidelines for AI procurement globally through the World Economic Forum’s network.

AI conference image

Collaborating with the World Economic Forum Centre for the Fourth Industrial Revolution

In practice, this means that as a World Economic Forum Fellow from the Office for AI, I’m working closely with a team in the Centre for the Fourth Industrial Revolution to establish best practice in AI procurement. We are learning from insights established by academia, industry, and governments around the world with the aim of testing the recommendations in the UK context over the autumn and then sharing the findings globally.

As the Forum’s Head of AI Kay Firth-Butterfield has said that:

“The future of AI needs government and businesses to work together. I’m thrilled to have the United Kingdom partner with the Centre on this project.”

The Centre is a newly established hub in San Francisco that brings together governments, leading companies, civil society and experts from around the world to co-design and pilot innovative approaches to policy-making and the governance of technology. Its vision is to shape the development and use of technology in ways that maximise the benefits and minimise the risks.

For example, we tested our approach at the World Economic Forum’s Annual Meeting of New Champions in Dalian, China, earlier this month with senior industry and government leaders around the world, which gathered interest from Bahrain to follow the UK’s lead and pilot the guidelines.

Dalian image

Representing a key private sector partner to this project, Lenny Stein Senior Vice President, Global Affairs from Splunk said:

“Splunk is privileged to be part of this important project and to work closely with the WEF and UK Government. AI can provide huge benefit to the citizen, provided that such technology is responsibly implemented with appropriate transparency, robust privacy protections, and assurance that the AI will achieve the intended outcome in a fair and ethical way. This project aims to help achieve this outcome.”

Get in touch

If you would like to receive more information or to get involved you can get in touch at AI@weforum.org

Share this