Policy plays a crucial role in influencing where, how and for whose benefit machine learning systems are developed and deployed. Safe and reliable deployment of machine learning systems requires policy frameworks that embed trustworthy data governance; that promote the use of machine learning in areas where it has potential to improve public wellbeing; and that account for the wider implications of technological change on individuals and communities. Research in this theme considers what policy levers can shape the development of AI technologies.
What policy levers can shape the development of AI technologies for societal benefit? ML@CL’s policy projects identify policy interventions that can support the development of trustworthy AI technologies and that help share the benefits of AI across society. Working with partners in civil society, national government, and international organisations, our work considers:
The AI Adoption programme studies how institutions take up AI in ways that deliver public value — through public dialogue, practitioner partnership, and emerging theory on judgement, model minimisation, and the information topography of organisational change.
The UK AI Council was an independent committee that provided advice on AI policy and strategy to the UK Government from 2018 – 2023. In 2022-23, the Council convened a series of discussions focused on the policy implications of advances in Large Language Models, with the aim of supporting rapid Government action to build national capability in Foundation Models.
The CDEI is a government expert body enabling the trustworthy use of data and AI. Its multidisciplinary team of specialists, with expertise in data and AI policy, public engagement, computational social science and software engineering, are supported by an advisory board of world-leading experts to deliver, test and refine trustworthy approaches to data and AI governance, working with organisations across the UK.
Reaping the benefits of data and digital technologies will require robust new institutions or frameworks that can allow data sharing - helping develop new data-enabled products and services - while protecting individual rights and freedoms. Data trusts offer a mechanism to achieve this goal through participatory data stewardship.
The COVID-19 pandemic intensified the need for timely and accurate data to inform policymaking. Countries across Africa have developed, and are still developing, innovative uses of data and statistics to monitor the extent of the spread of COVID-19 among their populations. Drawing from real-world examples, this projects the lessons offered by these innovations for future data policy frameworks.
European Learning and Intelligence Systems Excellence (ELISE) is a consortium of artificial intelligence (AI) research hubs that connects Europe’s leading researchers in machine learning and AI.
DELVE is a multi-disciplinary group, convened by the Royal Society, to support a data-driven approach to learning from the different approaches countries are taking to managing the pandemic. DELVE operated through 2020, providing advice to the UK Government and SAGE.
An independent advisory panel established by the Government Digital Service under DSIT to advise ministers and senior officials on the responsible use of AI across government. Neil is a panel member.
An independent expert advisory body established by the MHRA to recommend how the UK’s regulatory and assurance framework should evolve for the safe, effective and responsible use of AI in healthcare. Neil served as a commissioner and chaired the Technology Working Group.
The AI for Policy programme examines how AI changes the capabilities of public institutions and what is required to use those capabilities well. It considers both the opportunities created by AI and the institutional arrangements needed to preserve judgement, accountability, and democratic authority as parts of policy work become increasingly mediated by AI systems.
Our work examines how public dialogue can contribute to the democratic direction of AI and to better technological and institutional choices. It asks how societal priorities can be surfaced early enough to influence research and innovation; how public perspectives can inform decisions about AI adoption; and how institutions can develop the capacity to continue listening and responding as technologies and their uses change.
Africa Sandboxes Forum Lead, Datasphere Initiative
Project Manager, Data Trusts Initiative, Cambridge University
Co-Founder, Lichen AI
Director, ai@cam and Accelerate Science, Cambridge University
The DeepMind Professor of Machine Learning, Cambridge University
17 April 2020
The Royal Society launches DELVE — Data Evaluation and Learning for Viral Epidemics — to feed data-driven analysis into SAGE. Neil Lawrence helps lead the group; Jessica Montgomery provides secretariat.
4 May 2020
The first DELVE report argues that face coverings, including homemade cloth masks, could reduce community transmission when physical distancing is hard to maintain.
27 May 2020
A DELVE report sets out when TTI can help control COVID-19; Andrei Paleyes and colleagues release open simulation software used in the supporting analysis.
