Sweden's Migration Agency whistleblowers allege case officers lack basic Swedish, interpreters steer asylum outcomes

• Whistleblowers report case officers at Migrationsverket lack sufficient Swedish to conduct legally sound asylum interviews
• Interpreters — third parties with no formal vetting for bias — allegedly exercise outsized influence over interview outcomes
• Conversion cases singled out as particularly vulnerable to inconsistent and disputed assessments
• The agency denies all claims and insists its process meets legal standards

Whistleblowers with inside knowledge of Sweden's Migrationsverket (Migration Agency) are describing an asylum system where case officers sometimes lack the Swedish language proficiency needed to conduct reliable interviews, and where interpreters — third parties with no systematic vetting for conflicts of interest — can effectively determine whether an applicant receives protection or a deportation order. Samnytt reports that the allegations come from individuals with direct experience of the agency's internal processes, and that Migrationsverket has dismissed the claims, maintaining that its procedures are legally sound.

The specific complaints centre on three points. First, that some case officers responsible for asylum interviews do not command Swedish well enough to accurately capture nuance in applicants' accounts — a problem compounded when interviews are conducted through interpreters working in languages the officers cannot verify. Second, that assessments in so-called conversion cases — where asylum seekers claim to have converted to Christianity or another faith, making return to their home country dangerous — are inconsistent and sometimes driven by the individual officer's subjective judgment rather than any standardised methodology. Third, and most structurally significant, that interpreters occupy a position of enormous and largely unchecked power: they are the sole linguistic bridge between applicant and officer, yet face no rigorous screening for personal, political, or sectarian sympathies that might colour their translations.

This last point deserves particular attention. An interpreter who softens, sharpens, or subtly distorts an applicant's testimony can shift the entire evidentiary basis of a case — and neither the case officer nor the applicant may ever know it happened. The asylum interview is the single most important moment in the process. Swedish asylum law places heavy weight on the applicant's oral account and its perceived credibility. If the person translating that account is, in effect, an unaudited co-author, the system has a structural vulnerability at its most critical juncture.

Migrationsverket's blanket denial follows a familiar pattern. The agency has faced similar criticisms before — from lawyers, from former employees, and from Sweden's own parliamentary ombudsmen — and the standard response is that quality assurance mechanisms exist and that the process meets legal requirements. What the agency has never convincingly demonstrated is how it detects interpreter bias in real time, or how it ensures that officers who struggle with Swedish can nonetheless produce interview records that hold up to judicial scrutiny on appeal. The denial, in other words, addresses the accusation without addressing the mechanism.

Sweden processed tens of thousands of asylum cases during the peak years of 2015–2016 and has since scaled back dramatically, but the staffing and competency challenges that emerged during that period did not disappear with the caseload. The agency was forced to hire rapidly, and many of those hires came from backgrounds where Swedish was a second or third language. Whether the current workforce has been adequately retrained or replaced is a question Migrationsverket has not publicly answered with data.

The stakes are not abstract. A flawed interview can mean protection for someone who fabricated a conversion story, or deportation for someone whose genuine fear was lost in translation. Both outcomes corrode the system's legitimacy. The agency's quality controls are, by its own account, robust — but the controls are applied to a record that was already filtered through an interpreter no one was watching.

Sources: Samnytt