Abstract: Artificial intelligence and machine learning models have experienced a transformative evolution in capability and deployment. Recently, large-scale models trained on expansive datasets have revolutionized numerous domains, from online search to media creation. However, with the growing power of these AI systems, the urgency to address their security concerns has also escalated. Understanding and fortifying the robustness of such models is paramount in the pursuit of trustworthy machine learning. In this talk, I will delve into our research focused on uncovering the vulnerabilities of deep learning models in practical blackbox scenarios. I will shed light on our innovative methods for conducting efficient blackbox attacks, utilizing ensemble-based and context-aware strategies that demonstrate greater effectiveness than previously established techniques. Furthermore, the talk will explore potential future research avenues, emphasizing the need for continued vigilance and innovation in securing AI systems against emerging threats.