By Bhavya Johari


India’s Immigration and Foreigners (Exemption) Order, notified on 1 September 2025, entrenches religion-based exclusion into statutory form under the Immigration and Foreigners Act, 2025 (Act No. 13 of 2025). Paragraph 3(1)(e) of the Order limits exemption from passport and visa requirements to persons  belonging to a minority community in Afghanistan, Bangladesh and Pakistan, namely, Hindu, Sikh, Buddhist, Jain, Parsi and Christian, who entered India on or before 31 December 2024 due to religious persecution or fear of such persecution. This framework mirrors the Citizenship Amendment Act’s religion-based criteria, under which thousands had been granted citizenship. In contrast, the Order withholds comparable protection to those fleeing religious persecution from these same countries if they are Muslim refugees, including Rohingya from Myanmar present in India. By the government’s own estimates, around 40,000 Rohingya live illegally in India, underscoring how these religion-specific exemptions exclude a sizeable refugee population from protective frameworks.

The Order’s explicit religion-based stratification crystallises selective humanitarianism, a practice whereby states extend protection to certain groups while excluding others on political or religious grounds, as a binding statutory regime. In doing so, it contravenes customary non-refoulement obligations and exposes India to systematic breaches of binding international human rights law (IHRL) obligations. Judicial correction requires applying heightened scrutiny standards established by the European Court of Human Rights (ECtHR) in cases such as Chahal v. United Kingdom (1996) and Saadi v. Italy (2008), which require courts to independently assess the real risk of harm to deportees and place evidentiary burdens on states to demonstrate receiving countries’ capacity to protect returnees, with no balancing against national security concerns. Applied to India, this would require courts to verify Myanmar’s safety through an independent assessment of evidence from the UN High Commissioner for Refugees (UNHCR) and the UN Fact-Finding Mission rather than accepting government assurances.

Structuring Discrimination Through Legislative Consolidation

By converting an ad hoc, notification-based regime into a consolidated statutory framework, the Order entrenches religion-specific exemptions that explicitly discriminate against Muslim refugee populations. Paragraph 3(1)(e) operationalises this by codifying protection only for specified non-Muslim minorities from Afghanistan, Bangladesh, and Pakistan, while excluding similarly situated Muslim refugees and all refugees from Myanmar, including Rohingya. In doing so, the Order converts previously reversible 2015 executive notifications into a legally durable framework of exclusion that is more resistant to subsequent policy reversal. Before the Order, protection could have been extended to the Rohingya through executive notification; now, a statutory amendment is required.

In addition, Sections 16-20 of the Order expand the powers of the Foreigners Tribunals nationwide, granting quasi-judicial bodies the authority to order arrest and detain those from non-exempt Muslim minorities while shifting the evidentiary burden entirely onto individuals to prove that they are not foreigners. These expanded powers operate in a context where, as of September 2024, official figures record 676 Rohingya in immigration detention in India, 608 of whom had no court cases pending, constituting indefinite arbitrary detention. This transforms detention into the enforcement mechanism for selective humanitarianism: those excluded by the Order’s religion-based criteria face indefinite detention, while exempted religious minorities receive pathways to legal residency. Section 12’s authorisation of detention without judicial oversight codifies this practice into law, establishing an unmediated expansion of executive power that is incompatible with the rule of law.

Contravening Binding International Human Rights Obligations

India’s framework contravenes the interconnected IHRL obligations of liberty, movement, and non-discrimination. Article 9 of the International Covenant on Civil and Political Rights (ICCPR) protects against arbitrary detention and requires the availability of effective judicial review. The United Nations’ (UN) Human Rights Committee’s 2024 Concluding Observations on the fourth periodic report of India urged that immigration detention be used only as a measure of last resort, following individualised assessment. However, current practice falls far short of this requirement. Mass Rohingya detention in closed facilities proceeds without access to courts or meaningful status determination procedures. This amounts to an arbitrary deprivation of liberty contrary to Article 9, the impact of which is recognised to be devastating for already traumatized populations. Through visits to detention centres, Refugees International observed that Rohingya refugees face severe mental health deterioration and physical ailments, including temporary paralysis, with no access to mental health support or adequate medical care, while non-Muslim refugees from the same countries are not subjected to indefinite detention. Detention has thus become the primary enforcement mechanism, selectively deployed against those excluded by the Order’s religion-based criteria.

These conditions of confinement operate alongside additional restrictions on movement. Section 5 of the Order empowers the government to block exits on grounds of public interest that are not specified. In doing so, the provision lacks the necessary clarity and proportionality required under Article 12 of the ICCPR, which protects the right to liberty of movement. In practice, authorities deny exit permits to Rohingya approved for third-country resettlement through the UNHCR, leaving them unable to depart despite receiving offers from resettlement countries. Detained refugees are thus doubly constrained: confined without judicial remedy and prevented from leaving India solely through executive discretion.

These restrictions on liberty and movement facilitate refoulement, a practice that violates customary international law. Article 3 of the Convention against Torture prohibits transfer to a risk of torture or serious harm. This obligation has achieved customary status and binds India notwithstanding non-ratification. In its July 2024 early warning, the UN Committee on the Elimination of Racial Discrimination, called on India to refrain from forcibly returning Rohingya to Myanmar. Despite this, in May 2025, the United Nations expressed concern over reports that 40 Rohingya were abandoned at sea by Indian navy vessels. By returning them to Myanmar’s jurisdiction, these deportations expose refugees to a context that the UN’s Independent International Fact-Finding Mission on Myanmar, characterised as involving genocide, crimes against humanity, and a continuing risk of their recurrence. They constitute paradigmatic refoulement violations, revealing that India treats binding customary obligations as discretionary preferences subordinate to discriminatory security narratives.

The Order’s discriminatory design underlies these violations. Article 2 of the International Convention on the Elimination of All Forms of Racial Discrimination (CERD) obliges States to eliminate laws that create distinctions based on race, national origin or ethnicity. Read in conjunction with ICCPR Articles 2 and 26, this extends to religion-based distinctions that have discriminatory effects. The Order limits exemptions to non-Muslim minorities from Afghanistan, Bangladesh, and Pakistan. It withholds comparable protections to the Rohingya and other refugees, constructing a two-tier system of protection based on religion. No legitimate and proportionate objective justifies such differentiation. While the government has invoked security concerns for stricter controls, these justifications cannot override India’s binding IHRL obligations under ratified treaties, demonstrating an approach that treats international legal commitments as flexible political instruments rather than binding constraints on state action.

Regional Divergence: Bangladesh’s Contrasting Protective Approach

Bangladesh’s practice illustrates that robust refugee protection is a policy choice, not a function of treaty ratification. Bangladesh has not ratified the 1951 Refugee Convention, yet it hosts close to one million Rohingya refugees, recognising them as Forcibly Displaced Myanmar Nationals while facilitating large-scale humanitarian access. The 2025-26 Joint Response Plan mobilises $934.5 million through 113 partners for education, healthcare, and protection services. In practice, Rohingya in Bangladesh receive biometric identity cards indicating they are persons of concern who should be protected from forcible return, have access to humanitarian services, including healthcare and education, and are not subjected to mass arbitrary detention.

This cooperation is structured through Memoranda of Understanding with the UNHCR and other agencies. India has deliberately declined to replicate these institutional mechanisms: refusing to recognize UNHCR refugee cards and denying UNHCR access to detention facilities, thereby treating refugee recognition as a discretionary policy rather than establishing cooperative frameworks with UNHCR comparable to Bangladesh’s Memorandum on voluntary returns. While Bangladesh faces protracted displacement and serious protection gaps, its model offers de facto protection without systematic immigration detention or routine deportation to Myanmar. Bangladesh focuses on humanitarian assistance and biometric registration for service delivery, whereas India relies on long-term detention and active removal, including maritime pushbacks. This contrast shows that non-ratification of the Refugee Convention does not compel refoulement or discriminatory detention; it highlights India’s enforcement strategy as a policy choice rather than a legal inevitability.

Regional structures further accentuate the normative weight of this divergence. The South Asian Association for Regional Cooperation (SAARC) Charter excludes contentious bilateral issues from its deliberations, as outlined in Article X. This limitation has prevented the emergence of a regional refugee regime. UNHCR’s 1997 Model Law for South Asia was proposed in response to the region’s persistent non-ratification of the 1951 Refugee Convention and its Protocol, but the Model Law was never implemented. In this institutional vacuum, state practice becomes a key source of regional norms: Human Rights Watch recognizes Bangladesh’s adherence to non-refoulement, and its avoidance of mass detention establishes an essential precedent for compliance with customary refugee and human rights obligations even in the absence of binding regional or universal refugee treaties. India’s deviation from even this regional standard reveals that its discriminatory practices reflect deliberate policy choices prioritising religious exclusion over customary law adherence, not inevitable gaps from institutional incapacity.

Pathways for Judicial Reform Through ECtHR and Comparative Jurisprudence

European jurisprudence demonstrates how customary non-refoulement obligations can be judicially enforced with rigorous procedural safeguards. The foundational principles were established in Chahal v. United Kingdom (1996), which held that Article 3 of the European Convention on Human Rights (ECHR) prohibits return to a real risk of torture or inhuman treatment regardless of national security considerations. They were reaffirmed in Saadi v. Italy (2008) and remain consistently applied despite intensifying securitisation pressures. Over time, principles have evolved to address contemporary challenges in migration management, as seen in Hirsi Jamaa and Others v. Italy (2012), wherein the Court held that Italy violated Article 3 by intercepting migrants at sea and returning them to Libya without examining individual claims, establishing that non-refoulement applies extraterritorially even during maritime interdiction operations. This reasoning is directly instructive for India, where Rohingya deportations and reported maritime pushbacks similarly operate as mechanisms of externalised migration control that circumvent any meaningful individualised assessment of persecution risk.

The Court has refined these standards to address indirect refoulement and emergency exceptions. Ilias and Ahmed v. Hungary (2019) established that states removing asylum seekers to third countries must verify actual protection capacity and functional asylum procedures in the receiving state. This means that diplomatic assurances cannot be used as substitutes for examining systemic deficiencies. This principle directly contradicts India’s reliance on Myanmar’s willingness to accept deportees without verifying protection against persecution. Deportees from India arrive in Myanmar-controlled territory, where their current whereabouts and condition remain unknown, with families reporting that deportees are hiding from Myanmar’s military junta in regions contested by rebel groups and pro-military militia, where widespread distrust of the Rohingya makes their fate uncertain.

Recent jurisprudence has further reinforced the resilience of non-refoulement against emergency exceptions. M.K. and Others v. Poland (2020) and D.A. and Others v. Poland (2021) found that Poland violated Article 3 by conducting pushbacks at the Belarus border despite declared emergencies. This rejected arguments that national security or migration instrumentalisation justified suspending procedural safeguards. S.S. and Others v. Hungary (2023) held that unauthorised entry cannot absolve states of protection obligations. The Court’s most recent articulation in H.T. v. Germany and Greece (2024)  synthesises these principles into a clear standard. It found that Germany violated Article 3 by transferring an asylum seeker without obtaining individual guarantees regarding detention conditions or chain refoulement risks. The judgment held that transferring states bear responsibility for verifying actual protection, regardless of European Union membership or bilateral agreements, and requires concrete individual guarantees rather than systemic presumptions.     

This approach was also evident in Safi and Others v. Greece (2022), concerning the degrading body searches of shipwreck survivors and the inadequate investigation into the sinking, which held that resource constraints cannot justify inhumane treatment or procedural shortcuts that violate dignity. These principles have been operationalised domestically in comparative jurisdictions, demonstrating their practical application beyond the European regional system. The UK Supreme Court in R (AAA) v. Secretary of State (2023) unanimously ruled the Rwanda policy unlawful, finding that diplomatic assurances could not overcome evidence of systemic deficiencies, including 100% refusal rates for certain nationalities. The Court emphasised that non-refoulement constitutes customary international law that is binding through multiple instruments. Crucially, it placed the evidentiary burden on the government to demonstrate the receiving state’s capacity rather than requiring asylum seekers to prove risk. This methodology translates ECtHR principles into domestic judicial review in a non-ECHR jurisdiction.

India’s Supreme Court practice stands in stark contrast to this trajectory. In Mohammad Salimullah v. Union of India (2021), despite International Court of Justice findings on Rohingya genocide risks and evidence that Myanmar’s military coup intensified dangers, the Supreme Court refused complete interim relief. It also ordered that deportations could proceed, providing that the procedure prescribed for such deportation is followed. The Court held that the Rohingya could not invoke Articles 14 and 21 of the Indian Constitution, as rights against deportation, derived from Article 19(1)(e), are available only to citizens. This contradicted the Court’s precedent in NHRC v. Arunachal Pradesh (1996), which recognised that Articles 14 and 21 extend to all persons, and ignored the Delhi High Court’s holding in Dongh Lian Kham (2015) that non-refoulement forms part of the Article 21 protections.

Critically, the Court imposed no substantive protection requirement, no mandate to verify Myanmar’s safety or assess individual genocide risk, rendering the procedural condition insufficient given documented atrocities. The main writ petition remains pending before the Court, with no final judgment issued nearly four years after the April 2021 interim order. During this time, deportations have continued, and the 2025 Immigration Order has formalised  discriminatory policies into statutory law.

Conclusion

India’s 2025 Immigration Order represents the institutionalisation of state-sanctioned discrimination that weaponises protection frameworks to entrench religious exclusion while systematically enabling refoulement. India’s practice demonstrates that non-ratification can become a cover for deliberate non-compliance rather than a legal incapacity. This reveals how protection has become contingent upon state discretion, even though binding customary international obligations demand universal application. Judicial intervention must transcend procedural formalities to impose substantive non-refoulement protections. Without such oversight, selective humanitarianism risks becoming accepted practice under international law.

Bhavya Johari is a Lecturer at Jindal Global Law School, O.P. Jindal Global University, India. He earned his undergraduate law degree from NALSAR University of Law, Hyderabad, India, and holds an LL.M. from Melbourne Law School, University of Melbourne, Australia.


The views expressed in this article belong to the author/s and do not necessarily reflect those of the Refugee Law Initiative. We welcome comments and contributions to this blog – please comment below and see here for contribution guidelines.