With the approach of March 2026, the spotlight is once more focussed on pro-rights institutions of security. Security agencies, in particular, obtain powers that have leads that knows no bounds with a questionably weak oversight. That challenge can only be addressed through legal means, as institutional routines designate who or what can access power in the daily. A credible response has to have Rules, Evidence, Participation and Remedy.
This governing framework comprises the Rome Statute, international human rights law and domestic public-law guarantees. There are sources for that that provide standards, but there must be institutions to interpret, implement and review those standards. Key actors are those of the courts, legislatures, inspectors, prosecutors, police leaders, defence institutions and civil society. The coordination within that group needs to maintain the independence, and not confuse responsibility.
In international politics sociolisation focuses on the social production of security. Threats don't come with fixed meaning— they are instituted, behavior is coded, high-risk groups are identified and coercive instruments are made available—and the acceptable evidence that proves it is there is determined. Those classifications dictate budgeting, authority and legal standing, as well as public accounts. A rights-protective approach which looks also at institutional processes that determine what constitutes danger is therefore explored.
A stringent account of blame is added by international criminal law. The severity of violations is dependent on individual responsibility and evidence, context, and ties to actions and authority. Records should be retained, and the institutional design should have a way to retain records, before controversy starts, because otherwise log books, informal orders, etc. might not be kept and wouldn't be helpful at the time of subsequent review. It is important to document to defend affected communities as well as officials who may be subjected to unproven allegations.
One issue that is always a problem is secrecy. While there are some data that are legitimate operational information that needs protection, there is also not a magic bullet that can be used as a response to every question of legality, necessity, or harm. The various forms of publicly-reported statistics, specialist judges, parliamentary committees and cleared inspectors can give multiple layers of scrutiny. Suitable courts make it possible to keep delicate information secure at the same time retaining democratic information regarding scope and effect of any sort of coercive power.
The key question is command culture, formal safeguards are either practical restraints or not. Institutions give their message with respect to which conduct is positively rewarded through practices such as training, promotion, performance assessment, legal advice and disciplinary practice. A stop, arrest or intelligence data based measure – in isolation to a written policy – has little force or influence on operational success. Restraint, accurate reporting, de-escalation and cooperation in reviews as a positive attitude should be rewarded by balanced incentives.
Remedies—make a difference on behaviors, too. Time limits, disclosure and enforceable outcomes are not considered to be essential to trust, and if a complaint process is not independent, it will further erode trust. The use of investigation, explanation, correction, discipline, compensation, and reform of institutional practices constitute effective systems. Public summaries should document patterns and changes that have been made and finalized, without revealing business, protected personal data, etc.
Four key elements for the practical reform agenda: independence, necessity, proportionality review and enforceable remedies. Each element covers a distinct failure point, namely: before action is taking place there is the authorization process, during the process, the documentation process, after the action, there is the reviewing process, when harm results, there is remediation. Separation of steps is important because, if there is no record or reason earlier, conditioning on the subsequent steps will be less effective.
This evidence should consist of detention records, use of force records, classified review summaries, procurement trails, complaints and judicial outcomes, among other evidence types. These materials will enable comparisons between agencies, regions, groups and periods of time. There can be clearly identified patterns, disparities and delays, as well as repeated institutional failure, without compromising confidentiality.
Communities to be analysed for equity should be those that are affected by racial or political policing, border or emergency policing. Formality alone has no answer as to if and how burdens cluster, or whether and how review access varies by income, status, geography, disability, ethnicity, age or language. Distributional evidence must inform and shape design and to resource allocation.
The formal rules in institutional analysis should be related to incentives. Bureaucracy is responsive to workload, targets, budget constraints, norms of their profession, political pressure, and foreseen oversight. The force of reform takes place when those forces direct towards lawful and fair conduct. A safety device that is in a book may be lost to pressure of operation.
Any measurement has its own dangers. Socio-economic, financial and environmental outcomes can be easily reflected in indicators, which can foster strategic reporting. Quantitative data should, therefore, be used alongside qualitative data, independent verification and clear definitions. Why disaggregated reporting is important: national-level aggregate figures can mask high rates of failure locally or among groups of children and young people.
The legitimacy is also impacted by public communication. Using jargon can be a smokescreen for the true issues at stake in administration decisions, and heightened rhetoric can overstate the level of uncertainty to be faced. Objectives, evidence, trade-offs, affected rights and challenge available should be clearly explained. While there is always a limit, it is important to be honest about that and be confident; otherwise, that may be an empty statement.
Implementation requires resources. There is a need for staff and experts to review bodies, need to keep records at their disposal, need to have secure tenure in the review body and need to mandate response. Money, interpretation, suitable venues AND time also are needed for community participation. The absence of funding for rights and participation obligations can turn them into 'aspirational obligations'.
When it comes time for institutional reform, however, there's no such thing as a false choice between universal standards and context. Powers of protection (legality, equality, notice, review and remedy) are general. The use of those protections is effectively put into practice based upon the local knowledge. Durable design includes common protections and culture-, geography- and experience-based procedures.
A comparative analysis can be used to reinforce the design of institutions, but by not mechanically copying foreign institutions. A variety of different constitutional systems allocate the power of policing, prosecutions, intelligence and emergency powers around in different ways. Comparisons are: to look for function (copy label). Helpful questions include: What happens when you're not independent? What does it take for you to be able to share information? What would it take for you to be able to stop illegal activities from happening? What are repercussions if you don't listen to what you are suggesting?
There is a need to integrate the application of the professional art of the educator with legal responsibility. Decision under uncertainty, contradiction of orders and lack of information and pressure by the public, can all be tested in a scenario training. Supervisors are to focus on reasoning as well as on the outcome. Reflective practice is a way of making sense for the institutions to know when they've done something wrong and when they were negligent or intentional.
The private acquisition of technology and services that enhance coercive capabilities should be scrutinized, particularly for the proliferation of coercive capability outside of known command structures. Contents of contracts should detail legal boundaries, auditing access to data, who owns it, security standards and what is required to end a contract. Delegated tasks will continue to be the role of public bodies. The commercial confidentiality will not stay on top of the Constitutional accountability.
There should be a single point of accountability for the lawfulness and redress for multi-agency interventions. Using the term joint task force can create a situation where either no body is aware of the approval action, the information is still pending, or where complaints are simply not made to any one body. Written protocols should be prepared with the assignment of responsibility prior to deployment. Joint-yielding should never lead to a joint evasion.
The diversity of institutions and dissent in protected form are also key to prevention. There are various risks that can be identified by legal, analysts, community liaison and operational staff. Mechanisms to pass on concerns should exist at the internal level which connect them with senior review without risk of a repercussion on the career. [Corrected culture diminishes illegitimate injury and strategic bad fortune.
Achain governance should be defined by having responsibility at each and every stage. In order for courts, legislatures, inspection, prosecution, law enforcement, defence and civil society to efficiently cooperate, there must be processes of authorization, delivery, monitoring and remedy and they all must be traceable. Escalation plans and written protocols shall be put in place for overlapping mandates. If we're thinking about the people that need protection, ambiguity does not do that, ambiguity procreates institutional boundaries.
There needs to be an assessment of risk both through what is done and what is not done. Rights may be infringed if there is excessive intervention, but some forms of harm may remain unaddressed if there is late or partial (disjointed) intervention. Justifying the use of available evidence, and less restrictive alternatives should be evident in the decision records. This can be reviewed later for the purpose of evaluating the reasoning and not for the purpose of making unstated assumptions.
In independent research there can be a connecting link between the official reporting and the experience of the community. From a secure access to data, ethical oversight and publication rights, these serve as a means to begin testing institutional claims. The confidentiality and community authority should be protected in a partnership. There is a need for a formal response to findings.
Fewer illegal interventions, quicker complaint investigation, better disclosure and track record of respecting the ruling of the courts should be indicators of success. The scales that make up these measures are a mix of outcomes, procedure and institutional capability. The objectives must be visible, terms of reference must be clear, and when objectives are not met it is important to have a description about what went wrong and how to rectify the issues.
Legality, capacity, participation, transparency and correction are key elements of the strongest institutional model, as provided in the Rome Statute, international human rights law and domestic public-law safeguards, but the real protection is down to the way court, legislature, inspectors, prosecutors, police chiefs, defence institutions and civil society exercise authority. Examples of such evidence will include: actual records of detentions; use-of-force information; classified-review summaries; involvement with complaints; and judicial results. Thus, when security is effective, it can be legitimized by its associations with rights protection and when rights protection is effective, it can be legitimized by its associations with security.
