Blog by Mohammad Shehadeh


Consider the following: a practitioner describes a case in which the same individual, giving the same account, appears coherent in one hearing and inconsistent in another. The only difference is the interpreter. Yet the official record does not treat the interpreter as a variable. Instead, it records the inconsistency as if it were the applicant’s.

This is not an isolated incident, but the predictable outcome of a system that selects asylum interpreters through a cost-driven procurement process, without mandatory accreditation requirements. It then treats the resulting record as a neutral and accurate transcript of what was said.

This article draws on four primary sources gathered during an independent investigation into interpreter procurement failures in United Kingdom (UK) asylum proceedings: two immigration practitioners, a professional interpreter and a professional standards body. These were identified through direct outreach to immigration law firms and the National Register of Public Service Interpreters (NRPSI) and selected for their experience of interpretation in asylum proceedings. Despite their different roles, they all point to the same problem: asylum proceedings assume a level of accuracy that the existing system cannot reliably provide.

Asylum interpreters in the UK are typically sourced through a chain that runs from the Home Office (interior ministry) to a primary contractor, then to a sub-contracted agency, and finally to the individual interpreter. The asylum applicant has no role in this selection. One practitioner described the result plainly: the chain determines who is heard and who is ultimately marked as inconsistent.

When interpreters change across hearings—as they frequently do—the way testimony is recorded changes with them, even when the underlying account is identical. What recordings reveal, when they are available, is that accuracy often breaks down during interpretation. The Home Office, however, does not treat this as an interpretation problem, but as an inconsistency problem due to a lack of trust in claimants.

The structure of the asylum interview compounds this dynamic. A second practitioner described it as a process designed to test testimony for failure rather than to hear it. Discrepancies are sought, and the official record is constructed as material for future scrutiny across the life of a case.

In this context, the interpreter functions as a filter through which the official record is produced. The interpreter’s specialist knowledge—including familiarity with regional dialects, religious terminology and the specific language of persecution—determines what survives that filtering. Language fluency alone is insufficient. An interpreter who approximates a culturally specific term—finding no equivalent and reaching for the nearest available word—creates a version that becomes an inconsistency. In one case, a lawyer identified a mistranslation of the word bab in Arabic, which had been taken to mean “gate” when the correct meaning was “door”. In this context, this prolonged the proceeding and complicated an already difficult case.

What the interpreter’s own perspective adds is the mechanism for this inconsistency. A professional interpreter described the conditions under which asylum interpretation takes place: significant time pressure, complex testimony, no preparation material and an institutional expectation of speed. Under these conditions, the interpreter makes choices to omit and edit. They prioritise fluency over precision because precision at such speed is not possible.

The interpreter described this as producing errors that would constitute automatic failure in any professional examination. The system, however, does not record these choices. They said:

“Their duty is to interpret what is heard and understood… interpreting is not knowing two languages, it is understanding two languages.”

This is the core of the problem. The damage is invisible because the process that produces it is invisible. The record does not show what was said in the source language; it shows what the interpreter produced. The system then treats that output as a verbatim account of what was said.

A professional standards body confirmed the structural conditions that make this possible. The Diploma in Public Service Interpreting (DPSI), the main professional qualification for public service interpreters in the UK, functions in practice as a basic cognitive test rather than a test of specialist knowledge. More importantly, it is rarely enforced. Asylum proceedings routinely rely on agency interpreters who have not completed the DPSI or any equivalent qualification.

The system operates without mandatory accreditation. There is no requirement for who can act as an asylum interpreters. An agency can supply an interpreter with any level of qualification, or none. The applicant has no means of verifying this; their legal representative may not know. Yet the record produced by an unqualified interpreter and that produced by a qualified one are treated identically by the institution.

The system measures consistency in testimony that has already passed through a variable it does not track. An applicant whose account is assessed for inconsistency across multiple hearings may have given a fully consistent account in their first language at every stage. What varied was the interpreter—but the record does not show this.

This matters because inconsistency is a credibility finding. And credibility, in cases where corroborating evidence is by definition scarce, is often determinative. The question of whether an applicant’s account holds together cannot be answered without knowing whether the account recorded is the account given.

The minimal requirement is that the interpreter be treated as a variable in the evidential record, not as a channel that passes testimony through unchanged. This means accreditation requirements with genuine enforcement. It means applicants having a mechanism to raise concerns about interpreter quality. It means the official record reflecting interpreter identity and qualifications alongside the testimony it contains.

Until then, what the system produces is the record of a procurement chain.

Mohammad Shehadeh is a Year 10 student at St Ambrose College in Manchester. In November 2025, he gave a TEDxTrafford talk, “The words I had to leave behind” about how languages don’t neatly translate culture.


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.