<![CDATA[Multi-factor authentication (MFA) has become a critical security mechanism for protecting digital services; however, the effectiveness of MFA depends heavily on the usability and reliability of account recovery processes. Existing evaluations of MFA recovery deployments report alarmingly high failure rates, with approximately one in two users unable to successfully complete recovery tasks under testing conditions even before localization factors are considered. This study investigates how localization influences recovery usability and argues that translation can systematically amplify recovery failures through predictable and classifiable mechanisms. Common issues include text expansion that truncates recovery instructions, grammatical reordering that displaces embedded identifiers, mixed bidirectional text that reverses verification codes, and typography-related height changes that obscure keyboard-focusable controls. These failures are not caused by linguistic inaccuracies but rather by correctly translated content interacting with interfaces that were not designed to accommodate the structural characteristics of target languages. To address this challenge, the study proposes a two-instrument engineering framework grounded in the principles of WCAG 2.2 accessibility standards. The first component is a recovery message schema that models required fields, interaction dependencies, and interface invariants across all recovery states. The second component is a localization hazard model that captures causal relationships among translation phenomena, structural interface failures, and resulting user harm. Together, these instruments establish a verifiable standard for localization-safe recovery design. A recovery workflow is considered localization-safe only when all schema properties are preserved, interaction invariants remain satisfied, and every identified hazard class within supported locales is mitigated through design and verification activities. The findings highlight that reliable recovery experiences cannot be achieved through linguistic review alone. Instead, localization safety must be embedded within release governance, verification architecture, accessibility testing, and localization management practices to ensure secure, inclusive, and globally deployable authentication systems?]]>
