In engineering education, where hands-on problem-solving and technical proficiency especially in physics-based learning are critical, the role of artificial intelligence (AI) tools like ChatGPT remains debated; whether AI serves as a breakthrough innovation or presents new challenges. This study seeks to bridge that gap by examining the impact of ChatGPT on mechanical engineering students in a project-based course. It explores how students used AI tools to understand key concepts, support group collaboration, and improve coding and writing tasks. Using survey data from first-year students encouraged to integrate AI into their coursework, the research provides insights into the ethical and educational implications of AI in engineering education, considering both its benefits and challenges. The findings indicate that while ChatGPT was widely utilized for coding tasks such as MATLAB programming and enhancing conceptual understanding, its impact on group collaboration was modest. Ethical concerns, including the temptation to misuse AI, highlight the need for structured guidelines to ensure responsible AI usage. The study also identifies the necessity of verifying AI-generated outputs, as AI tools may produce inaccurate or misleading information, particularly in technical problem-solving. This paper offers recommendations for optimizing AI-assisted learning, fostering critical thinking, and adapting assessment practices to balance AI's educational benefits with academic integrity. These insights aim to guide educators and policymakers in effectively integrating AI into engineering and physics education while addressing its challenges to create a productive learning environment.