Post 10 in the Guest Blog Series Preserving Public Values in The Automated State by Malavika Raghavan, Alexandra Sinclair, Giulia Gentile and Orla Lynskey

This blog series examined the insights emerging from the CIVICA-funded DigiPublicValues project on the implications for public values when digital systems provided by private operators become embedded within the State. Key values of publicness, proportionality,
participation, transparency and accountability emerged through case studies on the use of digital technologies in education, border control and the administration of justice. Although these contexts differ significantly, common threads emerged from the research. In this final post, we offer a snapshot of the various contributions to this SLSA Blog special issue, followed by concluding remarks on common themes and future research directions.
DigiPublicValues: a Snapshot
DigiPublicValues has advanced the discussion on several interconnected themes concerning the protection of public values in the Automated State. To begin with, the umbrella of ‘Digital Constitutionalism’ offers a framework to think about public values in the digital environment, as explained by De Gregorio and Pollicino in their contribution to this special issue. Yet, as argued by De Abreu, the concept of ‘publicness’ is complex and cannot be reduced to a single, monolithic definition. A tension between ‘publicness’ and ‘privateness’ emerges in Breaugh’s and Hammershmid’s work, which showcases that without robust governance and public participation, public-private
partnerships built to digitise the State risk prioritising private interests over public good. Sanchez-Graells’ contribution gives further substance to these observations, and offers a study of the contractual mechanisms used in the public procurement of digital
technologies—and the need for new regulatory strategies to overcome the challenges they raise for public values such as transparency.
However, as noted in Curtin and Morandini’s contribution digitalising public services does not inevitably erode values of transparency and accountability. Rather highlighting these issues allows us to call for more careful thinking on how public-private
partnerships may be built for accountable digitalisation that extracts guarantees of transparency and accountability—including from private technology providers. Doing so, as Menendez and Syrrakos’ highlight, will require a reimagining of role of proportionality as a public value that acts as a bulwark against private and public actors’ abuse of administrative power. Similarly, Almada and Lopes rethink participation in public-private partnership to digitise public services, and advance the view that ‘public-in-the-loop’ should become embedded in those collaborations. While lawyers tend to look for law tools to govern the technology, Raghavan highlights the role of soft law as an instrument of governance, with its strengths and limitations.

Through the comparison of the case studies and the public values, the the DigiPublicValues project has thus shed light on transversal challenges from the angle of governance, regulation and established constitutional principles. Such challenges are the focus of future research from several members of the project.
Reflections on DigiPublicValues  

DigiPublicValues yielded some overarching insights. First, general principles such as transparency, accountability and participation are given effect to in different, specific forms in the legislation and case law that regulates digitalised systems. This begs the question of
whether it is necessary, or desirable, to study these principles holistically across fields of law, or to examine particular legal provisions and how they instantiate the broader public values of concern. Take for instance the AI Act’s requirement for an ‘appropriate type and degree of
transparency’ in the design and development of high-risk AI systems. Commentators have noted that this gives latitude to providers to determine what is appropriate information disclosure without necessarily providing details about the system’s inner workings. Likewise, the information requirements imposed on the processors under the GDPR are formulated as rights for the data subject, and have accordingly different, specific features that mirror the objectives of that legislation. Similar to these approaches, in English administrative law (which applies to public digital systems and procedures) there is no general duty of transparency. Instead, the extent of information disclosure is dependent on many factors including at which point of the decision-making process information is sought and the nature
of the interest at stake.

Beyond the formulation of the norms, or their translation into legal instruments, a second theme that emerged was about the variability of these norms when applied in different sectoral contexts. Certain uses of technology are permitted in some sectors but not in
others—which seems to reflect the extent to which a particular public value applies alongside competing considerations. For instance, transparency as a public value can be foregone in certain sectors, as seen in the UK where a data protection exemption exists for data
protection rights where they would likely prejudice the maintenance of effective immigration control. The EU’s AI Act also places limits transparency in areas of law enforcement, migration, asylum and border control. However, the same values may gain importance in other sectors depending on the political salience of the issue in question. Transparency and accountability obligations are heightened for instance, in relation to children’s rights as these are recognised differently by data protection law, especially relevant in educational contexts. This means that what may be permissible use of technology in one sector may be impermissible or require greater information disclosure or safeguards in another.
Further research is needed to make explicit the basis on which public values are upheld or eschewed within these sectoral-level formulations, and to analyse the legitimacy of these choices. At the same time, broader questions of governance, regulation and protection of established constitutional principles also emerge, and are at the core of the future research agenda of several members of the DigiPublicValues’ team.

For instance, Prof Orla Lynskey’s work interrogates what, if anything, is distinct about these constellations of public-private cooperation; what they mean for the continued legal relevance of the public-private divide and, more fundamentally, whether ‘publicness’ is
desirable per se and why.

Building on her research on the digitisation of courts in the UK and Italy under DigiPublicValues, Dr Giulia Gentile’s research addresses two questions: first, what is the impact of digitisation on the principle of separation of powers; second, how the digital evolution affects the social contract against the current historical background, characterised by a regression of liberal-democratic values.

Malavika Raghavan’s ongoing doctoral research unpacks the use of soft law as a mechanism that enables legal change to accommodate the demands of IT systems used in public sector digitalisation. It highlights the role of soft law as a vehicle for transporting the values and
priorities of technical systems into existing legal frameworks—in ways that challenge understandings of how legal change is initiated within the traditional legal hierarchy. Alexandra Sinclair’s research examines existing doctrines of judicial review and analyses how these doctrines in their current forms likely apply to uses of automation and machine learning in government. Her research examines to what extent these doctrines are capable of meeting their original objectives of facilitating legal redress for individuals and holding government to account in an environment of changing administrative behaviour due to uses of automation and machine learning.

We hope that ideas in this blog series, and in our forthcoming research, support the development of more nuanced, public-value-based and rights-protecting technological arrangements within semi-privatised systems of the Automated State.

Bios

Malavika Raghavan is a doctoral student at the LSE. Her work examines the interplay of law and technology within information systems in the public sector (specifically digital identity and welfare infrastructures) and their implications for the relationship of the State with
individuals.
Alexandra Sinclair is a doctoral student at the LSE. Her work explores the application of English judicial review doctrines to automated decision-making.

Giulia Gentile

Giulia Gentile is Lecturer in Law at Essex Law School. Her research focuses on the digitisation of the state, and the evolution of constitutional principles in the automated state.

Orla Lynskey

Orla Lynskey is Chair of Law and Technology at UCL Laws and a Visiting Professor at the College of Europe, Bruges. She is interested in the impact of private power on the protection of fundamental rights and the provision of public services.