Hong Kong Six Years After the National Security Law

1. The news facts: what has happened

This article examines the expansion of national-security governance into policing, education, public memory, professional life and ordinary social control in Hong Kong. Six years after the National Security Law, Human Rights Watch reported further expansion of the security system. Measures highlighted in 2026 included stronger powers to compel access to electronic devices, the possible use of national-security procedures in cases initially presented as ordinary crimes, additional confidential security funding and continuing arrests and convictions under the security framework.

The principal chronology is as follows. The relevant period begins with the 2020 National Security Law, continues through the 2024 Safeguarding National Security Ordinance, and extends into 2026 measures affecting police powers, public finance, education, museums, civil society, business and contact with overseas organisations.

Public reports must be read according to their actual scope. A questionnaire records the experiences of respondents; an official statistic describes the scale of a state process; and a case statement records the position of a particular person or organisation. None automatically proves every wider allegation, but repeated and consistent accounts create a serious duty to investigate.

Important facts remain unresolved: the full use of device-access powers, the number of ordinary cases shifted into security procedures, the breakdown of confidential spending, and the long-term effects on education, professional independence and public participation. These gaps should not be filled by speculation. Where state agencies hold the relevant decisions, detention logs, medical records or surveillance material, meaningful accountability requires that the information be made available for independent scrutiny.

2. The issues that deserve attention

The first issue is the boundary of public power. The relevant standards include legal certainty, freedom of expression and association, privacy, fair-trial guarantees, judicial independence, access to legal advice, public accountability and the requirement that security restrictions remain necessary and proportionate. The question is not whether a state may regulate conduct, investigate crime or protect public safety. It is whether the power is clearly defined, directed at a demonstrable risk, independently reviewable and accompanied by an effective remedy.

The second issue is timing. Access to a lawyer, family notification, independent healthcare, a written decision and a safe complaint process matter most at the beginning of state intervention. Safeguards introduced only after a confession, injury, prolonged disappearance or medical deterioration cannot fully repair the original harm.

The third issue is control of the evidence. Relevant materials include legislation, subsidiary rules, police guidance, government budget figures, court judgments, education materials, museum changes, contractual security clauses, arrest and conviction statistics and recorded restrictions on organisations and professionals. When the same authority accused of wrongdoing also controls the underlying records, a bare denial is insufficient. Privacy can be protected while still allowing an independent court, inspector or medical expert to examine the evidence.

The fourth issue is unequal exposure to harm. journalists, lawyers, teachers, students, artists, civil-society organisations, overseas activists, their relatives and ordinary residents whose work or communication touches politically sensitive subjects. A formally neutral rule can therefore have a much more severe effect on people who lack legal representation, medical resilience, financial resources or a safe family network.

The fifth issue is the distance between law on paper and protection in practice. Rights written in legislation have limited value if complaints trigger retaliation, courts do not examine the substantive reason for restrictions, lawyers cannot act independently, and officials are not held accountable after credible allegations.

3. Wider discussion: why the case matters beyond one person

These events can produce a chilling effect far beyond the number of people formally detained or convicted. Communities watch which conduct attracts official attention, whether relatives are punished, whether legal representation is possible and whether public appeals lead to further pressure. The result may be self-censorship, withdrawal from professional work, abandonment of worship or cultural activity, and reduced contact with people overseas.

A chilling effect is difficult to capture in a single statistic, but it changes the social environment. National security and administrative management cease to be exceptional interventions and become forces shaping ordinary choices. People alter where they travel, whom they meet, what they publish and whether they seek medical or legal help.

A fair analysis must also consider the principal alternative explanation. Hong Kong and central authorities state that the security laws restored stability and target only a small number of offenders. That claim must be examined not only through prosecution totals, but through the wider changes in education, speech, institutional independence and the willingness of people to participate in public life.

Proportionality is central. Even when a government pursues a legitimate aim, the chosen measure must be necessary, targeted and no more harmful than required. Indefinite surveillance, pressure on relatives, obstruction of urgent healthcare or punishment of peaceful advocacy cannot be justified merely by invoking security or administrative convenience in general terms.

The material also has a professional legal use. The evolving system is relevant to asylum, extradition and non-refoulement assessments involving political activity, overseas advocacy or security allegations. Courts should examine the practical operation of trial arrangements and monitoring, not assume that the use of formal legal procedures eliminates human-rights risk.

Foreign governments have responsibilities of their own. Police cooperation, extradition, asylum, university security, platform governance and consular processes can all become points at which cross-border pressure is enabled or resisted. Requests from foreign authorities should receive an effective human-rights review rather than being accepted because they appear formally regular.

Businesses, universities, religious bodies and professional organisations should protect member data, provide a safe channel for reporting interference and preserve relevant records. Institutional neutrality does not require ignoring credible threats; silence can transfer the cost of geopolitical or commercial caution to individuals who are least able to protect themselves.

The deeper issue is public confidence in law. A legal system commands authority not only because it can punish, but because decisions can be understood, challenged and corrected. When concepts continually expand, procedures remain closed and independent oversight cannot enter, formal legality may be experienced as political control rather than impartial justice.

4. CRM assessment and conclusion

CRM’s assessment is that the central concern is not an isolated administrative failure. The evidence concerning the expansion of national-security governance into policing, education, public memory, professional life and ordinary social control in Hong Kong points to a risk that is continuing, institutional and capable of affecting several rights at the same time. The combined impact on liberty, health, family, profession, expression and life abroad cannot be understood by examining each measure in isolation.

A rights-respecting system must be measured by practical indicators: whether people can foresee the legal consequences of their conduct; obtain a lawyer and independent healthcare from the earliest stage; challenge the factual basis of a decision; protect their relatives from collective pressure; and regain ordinary life when a formal process ends. Without those conditions, procedure can appear lawful while producing substantive injustice.

The necessary safeguards are concrete: narrowly defined offences, independent courts, full access to counsel, transparent public spending, protection for peaceful speech and association, an end to family pressure and extraterritorial intimidation, and review of measures that turn ordinary administration into security enforcement. They must be externally verifiable through published data, independent access, preserved records, protection for complainants and meaningful remedies when violations are established.

CRM rejects two equally damaging approaches: minimising credible human-rights risk for diplomatic or commercial convenience, and presenting an unverified allegation as a proven fact for political effect. The first abandons affected people; the second damages the credibility of evidence. Independent monitoring requires both disciplined attribution and a higher demand for disclosure from institutions that hold decisive information.

National security, public order and administrative efficiency are not unlimited reasons to displace human dignity. When peaceful speech, worship, cultural identity, legal representation, overseas advocacy or access to medical care can be penalised through broad political judgments, the damage extends beyond the immediate victim to society’s ability to resolve disagreement through law.

CRM will continue to compare official records, United Nations material, court decisions, professional investigations and verifiable case evidence on a common timeline. New judgments, official responses or independent evidence may change our assessment. On the present record, however, there are sufficient grounds to demand disclosure, an end to unnecessary restrictions, and effective access to lawyers, healthcare, complaint mechanisms and protection from retaliation.