Under the guidance of a professor, I researched how adversarial examples could be used to attack or evade machine learning-based intrusion detection systems. I designed and tested attacks using public tools, evaluated their impact on model accuracy, and presented findings to highlight vulnerabilities in AI-based security solutions.