By Yen-Rong Lai and Lili Song
In conversations about refugee protection in Asia, Taiwan rarely comes up. Yet Taiwan has a long history of hosting refugees and asylum seekers, including those fleeing the People’s Republic of China (PRC) since 1949, Indochinese refugees in the 1970s and 1980s and Hong Kong activists seeking refuge since 2019. Today, the largest group seeking protection in Taiwan are Myanmar nationals.
For decades, Taiwan’s handling of refugees and asylum seekers has operated in a state of legal limbo. Taiwan is not a party to the 1951 UN Refugee Convention or its 1967 Protocol – which together form the cornerstone of international refugee law – and its exclusion from the UN system since 1971 means that it cannot formally accede to those instruments, even if it wishes to do so. That being said, Taiwan can enact national laws consistent with the provisions of those instruments or voluntarily comply with them in practice, if it wants to.
Indeed, Taiwan has proactively incorporated major UN human rights treaties, such as the International Covenant on Civil and Political Rights, into its domestic legal system. Since 2005, Taiwan has repeatedly attempted to enact a Refugee Act. In 2022, the government launched its National Human Rights Action Plan, identifying refugee protection as a priority area and committed to developing a legal framework for asylum procedures. However, to date, Taiwan has not established a formalised, predictable framework for refugee protection.
As Taiwan frequently highlights its commitment to democracy and human rights, as well as its desire to contribute to the international rules-based order, it faces growing expectations – both domestically and internationally – to demonstrate how those values apply to people seeking refuge.
Recent Progress and Limits and Deficiencies
Notably, Taiwanese courts have gradually begun to reflect international refugee norms. For example, the Supreme Administrative Court has in recent years explicitly referred to the principle of non-refoulement and acknowledged that immigration decisions should be assessed in light of international refugee norms. This marks a significant shift from earlier cases, when courts were often reluctant to engage with refugee law arguments.
However, these commitments and principles have yet to be fully reflected in domestic immigration legislation. Taiwan’s Immigration Act contains no reference to refugees or asylum, let alone provisions on refugee status determination, non-refoulement or refugee rights. In 2023, during the amendment process for Taiwan’s Immigration Act, several legislators proposed incorporating protection for individuals facing persecution into the Act. However, the proposals were ultimately reduced to an ancillary resolution that called on the National Immigration Agency (NIA) to observe non-refoulement and coordinate basic support for asylum seekers, rather than create statutory protections.
The Taiwanese government has since continued to rely primarily on administrative and policy-based mechanisms. In April 2026, the NIA publicly committed to processing such cases through an administrative mechanism consisting of three stages: (1) administrative investigation, (2) consultation with external experts and scholars and (3) inter-ministerial discussion and review. Under the administrative mechanism, applicants whose claims are approved through the three-stage review process may be issued a provisional alien registration certificate under Article 23 of the Regulations Governing Visiting, Residency and Permanent Residency of Aliens, enabling them to lawfully remain in Taiwan.
However, the NIA review mechanism lacks procedural transparency and many of the safeguards associated with a statutory asylum system, resulting in significant uncertainty. The NIA originally stated that people with a need for international protection could submit an asylum application at local NIA service centres. In practice, the NIA did not publish any information or instructions regarding how people seeking asylum could apply, nor did it provide frontline officers with guidance on how to handle such cases. There is no established standard procedure for receiving asylum applications.
This has resulted in a highly problematic development in practice: access to the NIA mechanism appears to have become dependent on NGO intermediation. Cases accepted for review have generally been submitted with the assistance of civil society organisations. In one instance, an applicant who attempted to submit a protection claim independently was turned away, and the relevant NIA service centre subsequently indicated that such claims would only be accepted with NGO assistance. In effect, civil society organisations have been compelled to assume an informal gatekeeping role with regard to accessing the review mechanism, despite having no formal mandate to perform such a function, as applicants may otherwise be left without meaningful access to the procedure.
Furthermore, the assessment criteria have never been made public. Applicants are not guaranteed an opportunity to present their claims directly to decision-maker. In particular, individuals expressing fear of persecution or intent to seek asylum at airports or seaports have no formal channel to lodge a claim, creating a risk of return without any substantive assessment of their protection needs.
Under the NIA review mechanism, decisions are made by an inter-ministerial committee whose members are not required to possess expertise in refugee protection. There are no clear pathways for unsuccessful applicants to challenge negative NIA decisions through administrative appeals or judicial review. The NIA mechanism also lacks stable institutional resources, meaning that it could be modified or discontinued without legislative oversight.
During the often lengthy assessment process, applicants are generally unable to work lawfully in Taiwan, forcing many into the informal labour market where they face heightened risks of labour exploitation and immigration penalties. Notably, individuals in need of international protection who are apprehended for overstaying or unauthorised work may face deportation without a meaningful opportunity to present their protection claims.
The abovementioned shortcomings became evident in February 2026, when the government issued its first batch of decisions on applications submitted under the NIA mechanism. The applicants included former members of Myanmar’s National League for Democracy, political activists and individuals fleeing armed conflict and military airstrikes. None of the applicants in the first batch of Myanmar cases was granted protection, leaving them at risk of deportation. Most applications were rejected on the grounds that the applicants had failed to provide ‘concrete evidence’ of persecution, despite the authorities having never disclosed the criteria used to determine eligibility for protection. The NIA further stated that applicants who had previously overstayed or engaged in unauthorised employment were ineligible for protection. This criterion is inconsistent with the principle of non-refoulement, which applies irrespective of an individual’s immigration status and should not be denied solely on the basis of irregular entry or stay.
As such, in the absence of a clear statutory basis, refugee status determination in Taiwan is currently subject to significant administrative discretion without adequate safeguards.
Why Legislative Progress Has Stalled
National security is often cited as the primary reason for delaying refugee legislation. However, this rationale is difficult to reconcile with Taiwan’s existing legal framework. Those from the PRC and Hong Kong seeking protection may obtain legal status through separate legislation following government review, whereas foreign nationals from other countries – whose protection claims are generally not associated with the same national security considerations – continue to lack any comparable pathway to lawful status. This distinction is also reflected in the government’s approach to refugee legislation since 2009, which has generally sought to address protection claims from persons from the PRC, Hong Kong and Macau through separate legal regimes rather than the Refugee Act.
The continued absence of domestic refugee legislation therefore cannot be fully explained by national security concerns alone. Rather, it also reflects the limited political incentives to advance refugee protection. As refugees and asylum seekers have no electoral constituency in Taiwan, refugee protection has remained a low legislative priority.
Yet leaving these gaps unresolved not only creates a serious human rights protection deficit, but may also place the state at risk of facilitating rights violations by returning individuals to situations of persecution without adequate protection screening. At the same time, asylum seekers who cannot safely return may remain in prolonged legal and socio-economic insecurity, without access to basic rights and vulnerable to exploitation. Over time, such exclusion may also generate broader social and governance challenges for Taiwan.
Conclusion
True democratic resilience is built on the predictability of the rule of law. When Taiwan protects refugee rights through informal mechanisms rather than codified laws, refugee protection remains vulnerable to executive overreach and shifting political priorities.
A Refugee Act alone would not solve every issue – as questions about capacity, security screening and regional migration pressures would remain – but a structured, codified refugee law would provide greater legal certainty for asylum seekers and clearer guidance for government agencies. It would also strengthen national security by replacing an opaque, informal grey market of migration with clear vetting protocols, background checks and orderly institutional oversight.
Moreover, a codified refugee law would provide stronger alignment between Taiwan’s human rights commitments and its domestic legal system. For a society that has repeatedly expressed its commitment to democracy and human rights and its desire to contribute to the international rules-based order, establishing a robust, transparent asylum legal framework may become one of the most visible tests of that ambition.
Yen-Rong Lai: Refugee Rights Campaigner, Taiwan Association for Human Rights
Lili Song: Senior lecturer, Faculty of Law, Otago University (acknowledgement: part of the research I presented in the article was supported by an international visiting fellowship at the Taiwan Foundation for Democracy Feb-May 2026)
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.