Over the past few years, there has been much discussion regarding the potential of automated-decision making (‘ADM’) systems powered by mechanisms of computational intelligence such as machine learning or deep learning (commonly referred to as ‘Artificial Intelligence’ or ‘AI’). To date, such forms of (big) data analysis are most prominently relied on by the private sector, such as the search algorithms used by online search engines or the recommendation algorithms used by e-commerce and entertainment services platforms. These forms of data analysis in essence offer three main benefits, namely the speed and efficiency of decision-making as well as an ability to detect correlations that may be undetectable to the human brain.
The efficiency, speed and correlations offered by these forms of data analytics are also appealing in the public sector. Indeed, various products of computational learning are already being used in administrative processes and will likely become much more prominent in future years. Whereas these techniques offer important potential benefits, they have also been the cause of concern. Indeed, the use of ADM in administrative settings raises numerous important legal and ethical challenges. This paper introduces these new elements in the administrative toolbox and to survey related consequences, in particular possible implications for the principle of transparency.

Artificial Intelligence (AI) is an area of research and technology application that can have a significant impact on public policies and services in many ways. In just a few years, it is expected that the potential will exist to free up nearly one-third of public servants’ time, allowing them to shift from mundane tasks to high-value work. Governments can also use AI to design better policies and make better decisions, improve communication and engagement with citizens and residents, and improve the speed and quality of public services. While the potential benefits of AI are significant, attaining them is not an easy task. Government use of AI trails that of the private sector; the field is complex and has a steep learning curve; and the purpose of, and context within, government are unique and present a number of challenges.

Artificial intelligence (AI) is becoming a part of our daily lives at a fast pace, offering myriad benefits for society. At the same time, there is concern about the unpredictability and uncontrollability of AI. In response, legislators and scholars call for more transparency and explainability of AI. This article considers what it would mean to require transparency of AI. It advocates looking beyond the opaque concept of AI, focusing on the concrete risks and biases of its underlying technology: machine-learning algorithms. The article discusses the biases that algorithms may produce through the input data, the testing of the algorithm and the decision model. Any transparency requirement for algorithms should result in explanations of these biases that are both understandable for the prospective recipients, and technically feasible for producers. Before asking how much transparency the law should require from algorithms, we should therefore consider if the explanation that programmers could offer is useful in specific legal contexts.

Machine-learning algorithms are transforming large segments of the economy, underlying everything from product marketing by online retailers to personalized search engines, and from advanced medical imaging to the software in self-driving cars. As machine learning’s use has expanded across all facets of society, anxiety has emerged about the intrusion of algorithmic machines into facets of life previously dependent on human judgment. Alarm bells sounding over the diffusion of artificial intelligence throughout the private sector only portend greater anxiety about digital robots replacing humans in the governmental sphere. A few administrative agencies have already begun to adopt this technology, while others have the clear potential in the near-term to use algorithms to shape official decisions over both rulemaking and adjudication. It is no longer fanciful to envision a future in which government agencies could effectively make law by robot, a prospect that understandably conjures up dystopian images of individuals surrendering their liberty to the control of computerized overlords. Should society be alarmed by governmental use of machine learning applications? We examine this question by considering whether the use of robotic decision tools by government agencies can pass muster under core, time-honored doctrines of administrative and constitutional law. At first glance, the idea of algorithmic regulation might appear to offend one or more traditional doctrines, such as the nondelegation doctrine, procedural due process, equal protection, or principles of reason-giving and transparency. We conclude, however, that when machine-learning technology is properly understood, its use by government agencies can comfortably fit within these conventional legal parameters. We recognize, of course, that the legality of regulation by robot is only one criterion by which its use should be assessed. Obviously, agencies should not apply algorithms cavalierly, even if doing so might not run afoul of the law, and in some cases, safeguards may be needed for machine learning to satisfy broader, good-governance aspirations. Yet in contrast with the emerging alarmism, we resist any categorical dismissal of a future administrative state in which key decisions are guided by, and even at times made by, algorithmic automation. Instead, we urge that governmental reliance on machine learning should be approached with measured optimism over the potential benefits such technology can offer society by making government smarter and its decisions more efficient and just.

Recently both the Bank of England and the Financial Conduct Authority have carried out experiments using new digital technology for regulatory purposes. The idea is to replace rules written in natural legal language with computer code and to use artificial intelligence for regulatory purposes. This new way of designing public law is in line with the government’s vision for the UK to become a global leader in digital technology. It is also reflected in the FCA’s business plan. The article reviews the technology and the advantages and disadvantages of combining the technology with regulatory law. It then informs the discussion from a broader public law perspective. It analyses regulatory technology through criteria developed in the mainstream regulatory discourse. It contributes to that discourse by anticipating problems that will arise as the technology evolves. In addition, the hope is to assist the government in avoiding mistakes that have occurred in the past and creating a better system from the start.

Also in the digital age, markets work properly as long as consumers are well informed. What is peculiar of the digital age is that consumers have become very fragile, also because firms can extensively manipulate the information that they produce and distribute to markets. Antitrust authorities may find their way to prosecute business manipulative conduct, as some rulings suggest. However, the enforcement of antitrust law is subject to precise circumstances and requires cumbersome proceedings, especially when dominant firms are involved. Therefore, a simpler and more widespread intervention is needed. Although over the years traditional disclosure regulation has showed its limits, today algorithmic analysis gives room to more effective forms of disclosure regulation. Therefore, the paper maintains that both regulators and antitrust authorities can use these new forms of disclosure regulation to perform better their functions. Thanks to algorithmic analysis, (a) regulators can provide consumers with targeted co-regulated disclosures; (b) while competition authorities, using their advocacy powers, can provide trustworthy rankings and reviews about firms’ ability to comply with antitrust and consumer protection laws.

The next generation of e-commerce will be conducted by digital agents, based on algorithms that will not only make purchase recommendations, but will also predict what we want, make purchase decisions, negotiate and execute the transaction for the consumers, and even automatically form coalitions of buyers to enjoy better terms, thereby replacing human decision-making. Algorithmic consumers have the potential to change dramatically the way we conduct business, raising new conceptual and regulatory challenges.

This game-changing technological development has significant implications for regulation, which should be adjusted to a reality of consumers making their purchase decisions via algorithms. Despite this challenge, scholarship addressing commercial algorithms focused primarily on the use of algorithms by suppliers. This article seeks to fill this void. We first explore the technological advances which are shaping algorithmic consumers, and analyze how these advances affect the competitive dynamic in the market. Then we analyze the implications of such technological advances on regulation, identifying three main challenges.

Innovations in networked digital communications technologies, including the rise of ‘Big Data’, ubiquitous computing and cloud storage systems, may be giving rise to a new system of social ordering known as algorithmic regulation. Algorithmic regulation refers to decision-making systems that regulate a domain of activity in order to manage risk or alter behaviour through continual computational generation of knowledge by systematically collecting data (in real time on a continuous basis) emitted directly from numerous dynamic components pertaining to the regulated environment in order to identify and, if necessary, automatically refine (or prompt refinement of) the system’s operations to attain a pre-specified goal.

It provides a descriptive analysis of algorithmic regulation, classifying these decision-making systems as either reactive or pre-emptive, and offers a taxonomy that identifies 8 different forms of algorithmic regulation based on their configuration at each of the three stages of the cybernetic process: notably, at the level of standard setting (adaptive vs. fixed behavioural standards); information-gathering and monitoring (historic data vs. predictions based on inferred data) and at the level of sanction and behavioural change (automatic execution vs. recommender systems). It maps the contours of several emerging debates surrounding algorithmic regulation, drawing upon insights from regulatory governance studies, legal critiques, surveillance studies and critical data studies to highlight various concerns about the legitimacy of algorithmic regulation.