Muhammad is a Google DeepMind Scholar currently pursuing his Master’s in AI for Science at AIMS South Africa. He is passionate about solving the alignment problem, specifically focusing on the robustness of Large Language Models in multilingual and low-resource contexts. His research is centered on identifying temporal and linguistic vulnerabilities in frontier models, with the goal of building safety guardrails that are robust across diverse cultural and linguistic landscapes.