Consequence · Which legal relationships are changing?
The governance questions arriving before the institutions.
A day inside the system
You wake in Lahore, Nairobi, Jakarta or Manchester — it no longer matters much where. Overnight, a tax authority’s risk engine has scored your last filing; you will never be told your score, only its consequences. Your commute is priced by an algorithm; your bus route was redrawn by another. A tender you bid for is screened by procurement software before any official reads it. Your mother’s pension arrives — or does not — depending on whether a database agrees she is alive. That evening, a recommendation system quietly decides which candidates’ speeches you will see before an election.
Not one of these events looks like a legal event. There is no courtroom, no notice, no signature, no official signing an order. Yet every one of them is an exercise of power over a person — and every one raises the oldest questions law knows: by what authority? according to which standards? answerable to whom? correctable how?
The law has answers to these questions for decisions made by people. It mostly does not yet have answers for decisions made through systems. The six fields below are where KSC.JUSNREM works on building them — each opening with an ordinary scene, because the scenes are where the law will actually be needed.
01Field
Constitutional authority.
The scene. A ministry buys an “AI-powered” case-management system. The contract runs to three hundred pages; the pages about data protection are thorough. Nowhere does the contract say what error rate is acceptable, who may inspect the model, how a citizen gets reasons, or what happens to fifteen years of decision records if the vendor relationship ends. Procurement is where the constitution meets the purchase order — and, usually, loses. A tender that looks like an IT acquisition is in substance a delegation of public power to private code.
Constitutions allocate authority among institutions, impose limits and create pathways of accountability. Intelligent and computational systems do not abolish those arrangements, but they can make their practical operation harder to see.
A public institution may remain legally responsible while the design choices determining an outcome are made elsewhere. A private infrastructure may perform a function of deep public consequence without acquiring the obligations normally attached to public authority. An automated process may produce consistency while making the source of judgment obscure.
Future leadership should ask:
- Which institution possesses the lawful mandate?
- Which actor controls the operative system?
- Can the authorised institution understand and alter the process?
- Can an affected person identify the decision-maker?
- Can a court or reviewing body reconstruct what occurred?
- Does responsibility remain with the institution even when execution is delegated?
KSC.JUSNREM intends to develop frameworks that help institutions map the difference between formal authority and operational power.
02Field
Responsibility and remedy.
The scene. A trading agent, a delivery drone, a procurement bot: something acts, and harm follows. The owner says the system exceeded instructions. The developer says it behaved as designed. The operator says it was configured by the vendor. The vendor’s terms disclaim everything. The harm, meanwhile, is not distributed at all; it landed whole, on one person. The default outcome — blame the last human in the chain, usually the least powerful one — is both unjust and, worse, a design incentive to keep a person in the loop purely as a liability sponge.
Law connects action to consequence through responsibility. Intelligent systems can complicate this connection by distributing design, data, deployment, supervision and execution among different actors.
The difficulty is not solved by assigning personality to a machine or by placing every consequence on the nearest human operator. Both responses may ignore the actual architecture of control.
The governing question is architectural: who created the conditions of action, who possessed the capacity to intervene, who benefited, who owed a duty, who knew or should have known, and which institution can provide an effective remedy?
Future law will need forms of responsibility capable of following these relationships without creating zones in which harm has no respondent.
Trace the problem in one decision.
Hypothetical learning scenario
A person’s access to a public benefit is affected by an automated assessment.
Interactive — select each stage to open its legal question
-
Which law authorises the decision, and what purpose does it establish?
-
Who translated the mandate into criteria, priorities and exceptions?
-
Which information represents the person, and how can it be corrected?
-
What logic, model or rule affects the outcome, and who can examine it?
-
Which official or system communicates and gives effect to the decision?
-
Which right, interest, dignity or dependency is affected?
-
Who must give reasons, who can reconsider, and which forum can provide effective relief?
The decision is not located at one point. Legal responsibility must remain coherent across the whole chain.
03Field
Law, code and infrastructure.
The scene. Nothing happens. That is the point. The overloaded truck is never fined because the highway gate will not open for it; the unlicensed seller is never prosecuted because the platform will not list him; the payment is never reversed because it was never possible to attempt it. The rule was not applied to conduct; the rule was the environment. Law enforced through design is perfectly effective and perfectly unappealable — there is no violation, no proceeding, no forum, and therefore no place where a wrongly configured rule can ever be challenged. A mistake in a statute produces cases; a mistake in an architecture produces silence.
The relationship between law and technology cannot be expressed through one slogan.
| Relationship | What changes | Opportunity | Constitutional risk | Question for leadership |
|---|---|---|---|---|
| Law as Code | Legal rules may be represented in structured or machine-readable form. | Consistency and accessibility of legal rules can improve. | Formalisation always involves interpretation; the formal model can drift from authoritative law. | Who makes that interpretation, how is ambiguity represented, and how does the model remain connected to authoritative law? |
| Code as Law | Technical architecture can regulate conduct by determining what is possible, visible or permitted. | Architecture can give practical effect to protections immediately. | It may govern more immediately than a written rule, without publicity or contest. | How does such practical regulation become accountable? |
| Law as Infrastructure | Legal and institutional arrangements may become embedded in the systems through which identity, property, public services, markets or remedies operate. | Law can become continuously available where people actually act. | Hidden assumptions embedded in infrastructure become difficult to challenge. | How do embedded arrangements remain inspectable, correctable and open to lawful challenge? |
| Law as a Service | Legal capabilities may be delivered continuously through digital systems. | Accessibility of legal capability can improve. | Service design may reduce law to convenience while excluding professional judgment, public authority or procedural fairness. | What must service design preserve for the delivery to remain lawful and fair? |
Law as Code
What changes. Legal rules may be represented in structured or machine-readable form.
Opportunity. Consistency and accessibility of legal rules can improve.
Constitutional risk. Formalisation always involves interpretation; the formal model can drift from authoritative law.
Question for leadership. Who makes that interpretation, how is ambiguity represented, and how does the model remain connected to authoritative law?
Code as Law
What changes. Technical architecture can regulate conduct by determining what is possible, visible or permitted.
Opportunity. Architecture can give practical effect to protections immediately.
Constitutional risk. It may govern more immediately than a written rule, without publicity or contest.
Question for leadership. How does such practical regulation become accountable?
Law as Infrastructure
What changes. Legal and institutional arrangements may become embedded in the systems through which identity, property, public services, markets or remedies operate.
Opportunity. Law can become continuously available where people actually act.
Constitutional risk. Hidden assumptions embedded in infrastructure become difficult to challenge.
Question for leadership. How do embedded arrangements remain inspectable, correctable and open to lawful challenge?
Law as a Service
What changes. Legal capabilities may be delivered continuously through digital systems.
Opportunity. Accessibility of legal capability can improve.
Constitutional risk. Service design may reduce law to convenience while excluding professional judgment, public authority or procedural fairness.
Question for leadership. What must service design preserve for the delivery to remain lawful and fair?
These relationships are connected but not interchangeable. KSC.JUSNREM intends to examine the constitutional and institutional requirements of each.
04Field
Sovereignty and security.
The scene. A government’s tax records, court files and citizen communications run on infrastructure owned abroad, priced abroad, and switch-off-able abroad. Nothing is wrong. The service is excellent. The state has simply ceased to be able to say no to its supplier — and no one ever decided that; it accreted, renewal by renewal. By the time anyone notices, the exit costs are constitutional, not commercial.
Sovereignty is often described as supreme legal authority within a territory. Intelligent infrastructure raises a further question: can authority remain meaningful when the systems through which it is exercised are not understood, controlled or replaceable by the state?
Dependence is not automatically loss of sovereignty. States have always relied on external knowledge, trade and technology. The strategic issue is whether dependence removes meaningful choice, obscures public power or prevents an institution from fulfilling its constitutional responsibility.
Future leadership should understand:
- where critical data and decision capabilities reside;
- who controls technical standards and updates;
- whether systems can be inspected and contested;
- which functions require domestic institutional capacity;
- and how security measures themselves remain lawful and reviewable.
KSC.JUSNREM intends to connect these questions to legal architecture rather than treating sovereignty as a technological slogan.
05Field
Future justice.
The scene. The prosecution’s witness is a printout. The sub-postmistress says the shortfall never existed; the system says it did; the court — applying a presumption that computers operate correctly — believes the system. Twenty years later, Parliament legislates her innocence. As machine-generated evidence becomes the spine of civil, tax and criminal proceedings, the rules that admit and weigh it become decisive.
Computational systems may improve search, translation, case administration, access to information and the organisation of evidence. They may also reproduce inequality, hide assumptions or encourage institutions to mistake statistical confidence for legal judgment.
The essential question is not whether justice should use technology. It is which functions may be assisted, which responsibilities cannot be delegated and which safeguards must travel with every use.
Any future justice system should preserve:
- notice and intelligible reasons;
- equality of participation;
- the ability to challenge evidence and assumptions;
- an identifiable responsible authority;
- independent review;
- protection against automation bias;
- and access to a meaningful human remedy.
Algorithmic assistance may support justice. It cannot become a reason for justice to lose its human and constitutional address.
06Field
Intelligent public institutions.
The scene. A widow’s support payment stops. No official decided this; a data-match between two agencies flagged an inconsistency, and the system acted on the flag. Across town, a man cannot bury his father: the register lists the old man as already dead — a clerical error years ago, replicated since across every system that trusts the register. Each institution says the same thing: our records come from the source; take it up with the source. The source has a queue. When a single registry becomes the root of civic existence, an error in it is not a record problem but an existence problem.
Public institutions need information to govern. Better information can improve planning, coordination and service delivery. Yet an informationally powerful institution is not necessarily a legitimate one.
Development Informatics asks how information systems can improve human and institutional outcomes while remaining accountable to the people represented within them.
Future public systems should be designed around:
- public purpose rather than data accumulation;
- proportionate collection and use;
- understandable decision pathways;
- inclusion of local and institutional knowledge;
- security and continuity;
- correction of inaccurate information;
- and remedies for those excluded or harmed.
The goal is not an automated state. It is a more capable institution that remains recognisably public, lawful and human.
An agenda, not a catalogue of completed products.
These are fields of inquiry and institutional construction. Their presence on this website does not imply that every doctrine, framework or system already exists.
KSC.JUSNREM makes them public because future governance requires questions to be formulated before solutions can responsibly be claimed.