By Zhenbin Zuo, Lecturer at Essex Law School

On November 21st, 2023, an insightful presentation titled ‘Governance by Algorithms: from social norms to laws, to numbers and to code’ was delivered by Mr. Zhenbin Zuo, lecturer from Essex Law School, at the prestigious international conference of Computational ‘Law’ on Edge 2023. This conference, organised by the ERC Advanced Grant research project ‘Counting as a Human Being in the Era of Computational Law’ (COHUBICOL) in collaboration with the Journal of Cross-disciplinary Research in Computational Law (CRCL), stands at the forefront of exploring the evolving relationship between law and emerging technologies including AI.

Computational ‘Law’ on Edge 2023 is a world-leading symposium that converges scholars and experts from both domains of law and computer science. It serves as a dynamic platform for cross-disciplinary debate on the cutting-edge issues in the field of computational law, and focusing on the future of law in the age of AI and technological breakthroughs.

In his presentation, Mr. Zuo challenged the ‘legal singularity’ or substitution hypothesis , which suggests that AI and computer code might replace traditional legal processes based on texts, hermeneutics and natural language (text-driven law). Contrary to this view, Mr. Zuo proposed a novel ‘scaling and layering’ framework to better understand how code-based algorithms complement rather than supplant modern legal governance both in historical evolution and current practice. In the analysis this novel framework also helps reveal deeper nature/limitations of our current text-based legal systems, and asks how lawyers and/or scientists can improve the functioning of law in the challenges of new technologies like AI.

Mr. Zuo’s analysis reveals a hybrid governance model of our modern world, consisting of social norms, laws, numbers, and computer code, each representing increasing levels of formalization and abstraction. He proposes that computer code, when existing as a governance mode, can only achieve effective and legitimate functions by referring to frames in statistics/numbers and laws, in a similar way that legal governance have to rely on the wider framing effect of social norms-based governance at its sociological and normative core. The detailed argument is two-fold:

1. Scaling: Historically more formal/abstract governance modes evolved from less formal ones, enhancing the capacity of societies to govern larger territories and populations.
2. Layering: At any period of time (including current practices), these more formal modes of governance rely on and are influenced by the less formal ones, creating institutional path-dependencies with both positive and negative implications.

A key argument in Mr. Zuo’s presentation is the necessity for lawmakers to recognise the distinct nature and limitations of each governance mode in different spatio-temporal environments. He emphasised the importance of avoiding over-reliance on any single mode, particularly ‘code’ and ‘numbers,’ to prevent potential lock-in effects, ‘concept drifts’ or ‘model decay’, and ‘cold-start’ problems in designing institutions and polices. He briefly discussed the example of Zuboff’s ‘Uncontract’ chapter in Surveillance Capitalism (2018) which warns of Google’s potential to solely rely on smart contract and stop the car engine of e.g. a mother on her drive to see child in hospital; and how tax calculation algorithms need to adapt to new categories of employment such as ‘worker’ after the UK Supreme Court’s decision on Uber BV v Aslam [2021]. He also touched on how this framework can help us better understand China’s Social Credit Systems which adopt automation in courts, governments, train stations, airports and other spaces for debts enforcement, a paper Mr Zuo presented at the previous Computational ‘Law’ on Edge conference in 2022, and forthcoming in the journal CRCL.

The presentation was well-received, with Dr. Noura Al-Moubayed, Associate Professor in Computer Science from Durham University, providing insightful comments and the audience engaging in a lively Q&A session. Mr. Zuo responded to the various questions, further enriching this cross-disciplinary debate.

For those interested in delving deeper into the nuances of Mr. Zuo’s presentation and the conference, a recording is available here. Additionally, Mr. Zuo welcomes further inquiries and can be contacted at zhenbin.zuo@essex.ac.uk.