Spotlight on AI Governance in AI Safety Summit Week

Nov 6, 2023

This week the use of artificial intelligence (AI) is under scrutiny possibly as never before. This is the week of the UK government’s long-anticipated AI safety summit, complemented by an independent AI fringe programme ‘AI for Everyone’, with events taking place across London.

The increased interest in AI stretches across the globe, with last week seeing the UN announcement of an AI Advisory Body which will support the international community’s efforts to govern AI. Yesterday, American President Joe Biden issued an executive order on safe, secure and trustworthy AI, which looks ensure that AI’s potential is  harnessed, whilst its risks are managed. Furthermore, the EU is in the final stages of passing the world’s first laws on AI.

Data for Policy has been at the forefront of the debate into governance with, by or of the algorithms which underpin artificial intelligence since it posed the question ‘Government by Algorithm?’ at its 2017 conference. Its 2024 conference continues to expand the discussion, this time posing the question ‘Trustworthy governance with AI?’ With this question, we hope to raise debate into the emerging capabilities, use cases, and best practices enabling innovation that could contribute to improved governance with AI, as well identifying concerns that are being raised regarding these advancements in areas such as data, algorithms, privacy, security, fairness, and potential risks. The conference is currently open for contributions in the form of full papers, abstracts and panels, deadline 27 Nov.

In addition to fostering dialogue and debate through its global conferences, Data for Policy has a major role in capturing information from these. Its partnership with Data & Policy journal (published by Cambridge University Press and Assessment) provides focus and academic rigor to the growing community of researchers coalescing at the interface of policy and data. The journal has published many papers on AI and algorithmic governance since its inception in 2019. We have created a list of these, and would like to highlight in particular the recent papers from Bell, Nov and Stoyanovich (NYU) ‘Think About the Stakeholders First‘ and Tan and colleagues’ ‘Artificial intelligence and algorithmic decisions in fraud detection: An interpretive structural model

The community also hosts less formal channels of information dissemination, with a blog on Medium: our community approach to dialogue in this space is highlighted by contributions from Zeynep Engin (Data for Policy) and Leid Zejnilovic (Nova School of Business and Economics) and colleagues. There is also an extensive collection of conference presentations viewable on our YouTube channel, preprints and abstracts available on Zenodo. Instructive contributions to the blog include

For this landmark week, we have curated a special AI governance’ YouTube playlist. We’d like to highlight in particular keynote lectures from Gianluca Misuraca (Poytechnic University of Madrid) on ‘Governance ‘of, with and by’ AI: Unravelling the future of AI for government’, Philip Treleaven (UCL) and Karen Yeung (University of Birmingham).