A 2024 open-access book from Cambridge University Press warns that governments' growing reliance on algorithmic systems to interpret and enforce regulation is quietly undermining the rule of law - sometimes with deliberate intent, sometimes by simple negligence - and existing legal safeguards in the European Union are falling short.
The book, Algorithmic Rule By Law, maps out how technologies meant to improve public decision-making can instead produce systemic injustice, weaken accountability, and concentrate power in ways that erode core tenets of liberal democracy. The problem, the research argues, is not merely one of faulty code or bias in datasets - it is a structural transformation of law itself when it gets translated into machine-executable instructions.
The difference between rule of law and rule by law
The book's central distinction rests on two ways law can function. The rule of law, as it is understood in liberal democracies, constrains political power, applies equally to rulers and ruled, and fosters legal certainty. But it also relies on textual openness and discretion, enabling courts and civil servants to tailor general rules to particular situations in a thoughtful balancing of norms. "Applying the law is not a mechanical exercise," the research notes, "but requires a thoughtful balancing of norms."
By contrast, rule by law describes a context where law still exists but works as an instrument of oppression - either because it is applied in a rigid, legalistic way that ignores its underlying purpose, or because it is deliberately weaponized to dismantle checks on power. Both scenarios, whether negligent or purposeful, undercut the protection law is supposed to provide.
How algorithmic systems push toward rule by law
When public authorities embed legal and policy rules into algorithmic systems, something changes in the nature of legal application. "When laws and policies are transformed from text to code to make them machine-readable and executable, something can get 'lost in translation'," the book explains. "This transformation is not a mere technical exercise - though it is often treated as such - but a deeply normative one." The people who design and develop these systems - referred to as "coders" - gain significant influence over public decision-making, shaping how legal rules operate at scale before any individual case is ever reviewed.
Instead of trained civil servants exercising discretion case by case, decisions are influenced ex ante by a small group whose choices can embed rigid interpretations, disregard the law's ambiguity, and reduce the space for equitable correction. The opacity that often surrounds these systems also makes it harder for legislative and judicial branches to check executive power, deepening existing asymmetries. As a result, algorithmic regulation can produce equality-violating outcomes across entire populations while making the discrimination harder to detect.
The five pathologies of algorithmic rule by law
The research identifies five interlocking risks that convert well-intentioned algorithmic regulation into a vehicle for rule by law:
- Primacy of techno-rationality: Efficiency and procedural rationality get prioritized over normative values such as human rights and administrative justice.
- Rule of coders: Power shifts from civil servants bound by deontological duties to the handful of people who design the systems, centralizing discretion in a way that pre-empts case-by-case judgment.
- Automation of legalism: The law's inherent openness is narrowed into a formalized, rigid application, stripping away the flexibility that allows the law to correct its own hard edges.
- Deficit of accountability: Opacity in design and implementation weakens oversight, making it harder to maintain constitutional checks and balances.
- Systemic vulnerability: The underlying technical infrastructure can be altered with a few clicks, enabling systemic breaches that are difficult to trace and impossible to challenge through individual cases.
Not every use of algorithmic regulation leads to these outcomes, the book stresses. But the conditions for harm are present and have already materialized. Two real-world patterns stand out. Governments with good intentions may scale a flawed legalistic policy - as the Dutch Childcare Benefits Scandal did - and in the process build infrastructure that a future, less benevolent administration can exploit. Alternatively, an autocratic-leaning government can roll out algorithmic systems under the guise of efficiency while using their opacity to push through illiberal policies, such as terminating grants related to equity and diversity or deploying facial recognition to police public gatherings.
For legal professionals monitoring the intersection of technology and governance, understanding these risks is critical. Resources on AI for Legal can help practitioners grasp how algorithmic systems alter the rule-of-law landscape. Meanwhile, the public sector's accelerating adoption of these tools makes AI for Government a relevant lens for assessing where oversight must tighten.
Existing legal frameworks remain insufficient, the research argues, even in jurisdictions with extensive regulation like the EU. While instruments such as the General Data Protection Regulation, the Conditionality Regulation, and the new Artificial Intelligence Act provide partial safeguards, significant gaps persist. The book calls for a mix of legal, organizational, and technical measures - public participation in algorithmization decisions, upstream oversight of development choices, and stronger constitutional checks - but acknowledges the difficulty of enacting them amid the geopolitical rush to adopt AI.
Why this matters for legal professionals
Lawyers, judges, and policymakers who work with administrative law, constitutional law, or technology regulation must now confront a reality where algorithmic systems do not merely automate existing rules - they reshape what legal interpretation and enforcement actually mean. The pathology of "rule by law" becomes harder to spot when it wears the appearance of neutral, efficient code. Recognizing the difference between a mechanical application that undercuts legal protection and a thoughtful balancing that upholds it is no longer an abstract concern; it is a practical demand for anyone involved in reviewing, challenging, or designing public-sector technology. The book's warning is that without deliberate intervention, the infrastructure for algorithmic rule by law is being built now, often with the best of intentions, and legal professionals will be the ones left to contend with its consequences.
Your membership also unlocks:
