LLMMedium
AI translation is jeopardizing Afghan asylum claims
- Occurred
- Jun 20, 2020
- Published
- Apr 19, 2023
- Source
- AI Incident Database
Summary
A Pashto-speaking refugee's asylum claim was denied by a US agency for a discrepancy between oral and written recount of an event allegedly due to an error of their automated translation tool which swapped pronouns of her written statement from "I" to "we".
Developers
- Unknown
Deployers
- US Citizenship and Immigration Services
Harmed parties
- anonymous Pashto-speaking refugee
- Pashto-speaking asylum seekers
- Dari-speaking asylum seekers
Classification
Category confidence: 90%
Severity confidence: 85%
Rationale: Automated translation error caused denial of asylum claim, representing a notable failure with real-world harm.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8