One of the primary goals of introductory programming courses is to develop fundamental programming competencies, including the ability to design and implement computer programs. Computer-based laboratory examinations, or lab exams, are considered a valid, efficient, and reliable assessment method for these competencies. In a lab exam, students are required to complete programming tasks in a proctored and time-limited setting with measures to prevent academic dishonesty. However, the proliferation of Generative Artificial Intelligence (GenAI) has raised new challenges for academic integrity during lab exams. In this work, we propose an approach to selectively block common GenAI tools during lab exams and share our preliminary experience using it in an introductory programming course.