The SAFER Act: A House Bill to Strip Asylum Over Return Travel — and the Debate Over Fraud, Due Process, and International Law

In early April 2026, Rep. Tom Tiffany (R-WI) introduced the Stopping Asylum Fraudsters Enforcement and Removal Act — the SAFER Act — a bill that would prohibit the Secretary of Homeland Security and the Attorney General from granting asylum to anyone who has returned to the country they claimed to be fleeing [1]. For those already granted asylum, the bill would allow the government to terminate their status and pursue denaturalization if they voluntarily traveled back while living in the United States [1].

"If someone claims they are fleeing danger and seeking asylum in the U.S., they should not be turning around and vacationing in the very country they said they had to escape," Tiffany said when announcing the legislation [1].

The bill permits an exception only if the State Department certifies that a "legitimate government transfer" has occurred and the original persecution threat is resolved [1]. It also addresses stateless migrants by analyzing their claims based on their most recent "habitual residence" [1].

The Case That Sparked the Bill

The SAFER Act's introduction followed the April 3, 2026, arrest of Hamideh Soleimani Afshar and her daughter Sarinasadat Hosseiny — the niece and grandniece of Qasem Soleimani, the late commander of Iran's Quds Force killed in a U.S. drone strike in January 2020 [2]. The Department of Homeland Security called Soleimani Afshar's 2019 asylum claim "fraudulent," citing at least four trips back to Iran after she received a green card in 2021 [2]. She had entered the United States on a tourist visa in 2015 and disclosed the return trips on a 2025 naturalization application [2].

The State Department said Soleimani Afshar "is an outspoken supporter of the totalitarian, terrorist regime in Iran," noting she promoted Iranian regime propaganda and celebrated attacks against U.S. forces while living a "lavish lifestyle in Los Angeles" [2]. Secretary of State Marco Rubio also revoked the residency of Fatemeh Ardeshir-Larijani, the daughter of a former Iranian security official, and her husband, who had already left the country [2].

The case offered Tiffany a clear-cut illustration of what he calls a systemic problem: people receiving protection from a country they demonstrably do not fear.

How Many Asylum Holders Travel Back?

There is no comprehensive federal dataset tracking how many of the roughly 54,350 people granted asylum annually (the FY2023 figure) return to their countries of origin [3]. USCIS treats return travel as a fraud indicator in individual case reviews but does not publish aggregate statistics [4]. The Government Accountability Office noted in a 2015 report that neither USCIS nor the Executive Office for Immigration Review (EOIR) had "assessed fraud risks across the asylum process" [5].

What data does exist suggests that confirmed fraud is rare relative to the volume of asylum grants. Between fiscal years 2010 and 2014, USCIS terminated asylum claims due to fraud in just 103 cases in 2010, declining to 34 in 2014 — against 76,122 asylum grants during the same period [6]. EOIR opened only seven fraud investigations in 2014, when immigration judges granted 8,775 cases [6].

Source: DHS Office of Homeland Security Statistics

Data as of Oct 2, 2024CSV

Immigration attorney Jason Dzubow, who runs The Asylumist legal blog, has documented the range of reasons asylum holders return: visiting critically ill or dying parents, attending funerals, managing inherited property, retrieving children left behind, and in some cases engaging in democracy and human rights work [7]. "Many of my clients face long-term threats in their countries," Dzubow wrote, but "some can return briefly...and remain relatively safe" — particularly those fleeing non-state actors like the Taliban rather than government persecution [7].

Existing Legal Tools: What the Law Already Allows

Critics of the SAFER Act argue that federal immigration law already provides mechanisms to address the concern Tiffany raises. Under current statute, USCIS can terminate asylum if a person has "voluntarily availed himself or herself of the protection of the country of nationality or last habitual residence by returning to such country" [7]. Immigration judges can also apply the "firm resettlement" bar — defined at 8 C.F.R. § 208.15 — to deny asylum to applicants who had permanent resettlement elsewhere before arriving in the United States [8].

The "changed country conditions" framework allows the government to reopen and revoke asylum when conditions in the home country have materially improved, and return travel can serve as evidence in that determination [9].

Proponents of the SAFER Act counter that these tools are underused and structurally inadequate. The existing standard requires the government to prove voluntary availment on a case-by-case basis, placing a resource-intensive burden on an already overwhelmed system. With 3.32 million pending immigration court cases as of February 2026 — approximately 70% of which involve asylum seekers — and only about 570 active immigration judges, the capacity to individually adjudicate return-travel questions is severely constrained [10].

Source: TRAC Immigration Reports

Data as of Feb 28, 2026CSV

Average wait times nationally have stretched to nearly 900 days from initial filing to final disposition [10]. The legal representation rate in immigration court stood at just 33.3% in February 2026 [10].

The International Law Question

The United States acceded to the 1967 Protocol Relating to the Status of Refugees, which incorporates the core protections of the 1951 Refugee Convention [11]. Article 1(C) of the Convention specifies that refugee status ceases when a person "has voluntarily re-availed himself of the protection of the country of his nationality" — language that mirrors the existing U.S. statutory standard [11].

However, the Convention's cessation clauses have been interpreted by UNHCR and international courts as requiring an individualized assessment of whether re-availment was truly voluntary and whether the underlying persecution risk has ended [11]. A blanket rule that automatically bars or terminates asylum based on any return travel — without an individualized inquiry — would, according to immigration law scholars, conflict with this framework.

Congress does have the constitutional authority to pass domestic legislation that conflicts with treaty obligations. Under the "last in time" rule established in cases like Whitney v. Robertson (1888), a later federal statute supersedes an earlier treaty as a matter of domestic law [12]. But such an action would place the United States in breach of its international commitments — a political and diplomatic consideration rather than a strict legal bar.

How Other Countries Handle Return Travel

Germany, Canada, and Australia each address the question of asylum holders traveling to their home countries, with varying degrees of strictness.

Germany requires the Federal Office for Migration and Refugees (BAMF) to review protection status when a refugee travels to their country of origin or contacts their home country's embassy [13]. But the German system conducts individualized assessments: authorities investigate the reasons for travel, whether it was voluntary, and whether the underlying threat persists [13]. In 2024, Germany initiated 17,578 revocation procedures but actually revoked status for only 2,229 persons — a revocation rate of roughly 12.7% of initiated procedures, and a fraction of the overall refugee population [14]. German case law also recognizes that "trauma or mental disorders that result from persecution constitute compelling reasons" that may justify continued protection even after a return visit [14].

Source: BAMF / AIDA

Data as of Jan 1, 2025CSV

Canada prohibits refugees from using their Refugee Travel Document to travel to the country of claimed persecution [15]. Doing so can trigger a review of status. However, the Canadian system does not impose automatic termination; instead, it evaluates the circumstances of travel.

Australia adopted a more rigid approach in 2013, applying Condition 8559 to protection visas, which prohibits travel to the country of origin. If such travel occurs, the visa can be cancelled [16]. The Refugee Council of Australia has warned that protection visas have been cancelled while refugees were overseas, sometimes with only seven days to respond to a cancellation notice and no right to appeal if the deadline is missed [16].

Source: UNHCR Population Data

Data as of Dec 31, 2025CSV

Due-Process Concerns and the Exceptions Gap

The SAFER Act's sole exception — requiring a State Department certification that a "legitimate government transfer" has resolved the original threat — is narrow by design. It does not explicitly address family emergencies, funerals, retrieval of minor children, or gathering evidence for legal proceedings [1].

Immigration advocates argue this creates a due-process gap. A Guatemalan mother who crossed the border fleeing gang violence and later returned briefly to retrieve a child she could not bring on the initial journey would, under the bill's text, face automatic disqualification from asylum [7]. An Afghan who received asylum based on Taliban threats and returned to visit a dying father would face the same outcome [7].

The bill's proponents might respond that such cases represent exactly the kind of individualized assessments that the current system already fails to conduct effectively. The question is whether the solution is to eliminate the assessment altogether or to fund the capacity to perform it.

With immigration courts processing cases at a deportation rate of 79.6% as of February 2026 and a bond grant rate of just 27.8%, critics question whether the system has the institutional capacity to fairly adjudicate individual exceptions even if they were written into the bill [10].

The Fraud Problem: How Big Is It?

The core factual dispute underlying the SAFER Act is whether asylum fraud is a marginal problem or a systemic one.

The available federal data points toward low confirmed fraud rates. The GAO's 2015 finding of 34 fraud terminations against tens of thousands of annual grants is the most frequently cited figure [5][6]. A University of Pennsylvania Law Review study found that 95% of asylum applicants attended every scheduled court hearing between 2008 and 2018, and 99% of those released with a Notice to Appear in FY2019 attended all hearings [6].

But fraud skeptics face a challenge: the government has acknowledged that its fraud detection tools are limited. The GAO specifically found that USCIS and EOIR have "limited capabilities to detect asylum fraud" and that "neither agency has assessed fraud risks across the asylum process" [5]. A 2014 New York investigation charged 30 defendants in asylum fraud schemes connected to 829 USCIS asylum grants and 3,709 EOIR grants — suggesting that organized fraud rings, while not the norm, can operate at significant scale when they exist [5].

The Center for Immigration Studies, which advocates for lower immigration levels, has argued that the credible fear screening process — the first step for many asylum seekers — has a pass rate that is itself evidence of insufficient scrutiny, though immigration courts ultimately deny a substantial share of those claims on the merits [17]. The Niskanen Center has countered that high denial rates reflect the narrowness of asylum's legal definition and procedural barriers — particularly the lack of legal representation — rather than fraud [18].

Unrepresented asylum applicants face denial rates around 90%, compared to roughly 50% for those with attorneys — a gap that immigration scholars attribute primarily to the complexity of asylum law rather than the merits of the underlying claims [18].

What the Bill Would and Would Not Do

The SAFER Act targets a real phenomenon — there are documented cases of individuals receiving asylum and then traveling freely to the country they claimed to fear. The Soleimani Afshar case is a vivid example [2].

But the bill's mechanism — a blanket bar with a single, narrow exception — differs fundamentally from the individualized approach used by every comparable asylum system. Germany reviews hundreds of thousands of cases and revokes a small fraction after investigation [14]. Australia's stricter approach has drawn criticism from its own Refugee Council for producing unjust outcomes [16]. Canada prohibits the travel but evaluates cases individually before revoking status [15].

The bill does not include provisions for additional immigration judges, fraud investigators, or administrative capacity. It does not address the 3.32 million case backlog that constrains the system's ability to distinguish legitimate from fraudulent claims [10]. And it does not include carve-outs for family emergencies, evidence-gathering, or situations where the asylum holder traveled under coercion.

Whether the SAFER Act represents a proportionate response to a genuine enforcement gap — or an overcorrection that would sweep up legitimate refugees alongside fraudsters — depends in large part on empirical questions that the federal government has not yet answered: how many asylum holders travel back, why, and what share of those trips indicate genuine fraud rather than the complicated realities of displacement.