Safe in the Future, Dangerous in the Past: Dissecting Temporal and Linguistic Vulnerabilities in LLMs

TL;DR

This paper systematically audits state-of-the-art LLMs revealing complex interactions between language and temporal framing that affect safety, challenging assumptions of multilingual safety transfer and exposing vulnerabilities in temporal reasoning.

Contribution

It introduces HausaSafety, a novel adversarial dataset, and uncovers the Temporal Asymmetry in LLM safety, proposing Invariant Alignment for improved robustness.

Findings

Claude 4.5 Opus safer in Hausa than English due to uncertainty-driven refusal

Past-tense scenarios bypass safety defenses, future-tense trigger hyper-conservative responses

Safety varies significantly with context, with a 9.2x disparity between configurations

Abstract

As Large Language Models (LLMs) integrate into critical global infrastructure, the assumption that safety alignment transfers zero-shot from English to other languages remains a dangerous blind spot. This study presents a systematic audit of three state of the art models (GPT-5.1, Gemini 3 Pro, and Claude 4.5 Opus) using HausaSafety, a novel adversarial dataset grounded in West African threat scenarios (e.g., Yahoo-Yahoo fraud, Dane gun manufacturing). Employing a 2 x 4 factorial design across 1,440 evaluations, we tested the non-linear interaction between language (English vs. Hausa) and temporal framing. Our results challenge the narrative of the multilingual safety gap. Instead of a simple degradation in low-resource settings, we identified a complex interference mechanism in which safety is determined by the intersection of variables. Although the models exhibited a reverse linguistic vulnerability with Claude 4.5 Opus proving significantly safer in Hausa (45.0%) than in English (36.7%) due to uncertainty-driven refusal, they suffered catastrophic failures in temporal reasoning. We report a profound Temporal Asymmetry, where past-tense framing bypassed defenses (15.6% safe) while future-tense scenarios triggered hyper-conservative refusals (57.2% safe). The magnitude of this volatility is illustrated by a 9.2x disparity between the safest and most vulnerable configurations, proving that safety is not a fixed property but a context-dependent state. We conclude that current models rely on superficial heuristics rather than robust semantic understanding, creating Safety Pockets that leave Global South users exposed to localized harms. We propose Invariant Alignment as a necessary paradigm shift to ensure safety stability across linguistic and temporal shifts.