The development of artificial intelligence technology has made significant contributions in various fields, including education. One problem still frequently encountered in higher education is the manual student attendance system. This manual process often creates various obstacles, such as the potential for fraud in the form of absenteeism, delays in recording, and the risk of administrative errors. This condition shows the urgency of providing a more modern, effective, and accurate attendance system through the use of AI-based technology. Using the YOLOv11 algorithm, known for its ability to detect objects quickly and with high precision, this research is directed at designing a Smart Attendance system based on facial detection that can improve the quality of academic services at Muhammadiyah Ahmad Dahlan University, Palembang. The purpose of this research is to develop and implement an automatic attendance system based on facial detection with the YOLOv11 algorithm integrated with the campus academic database. This system is expected to be able to recognize student faces in real-time, record attendance automatically, and present data that can be accessed by lecturers and administration in a transparent manner. Furthermore, this research aims to improve the efficiency of academic processes, reduce attendance fraud, and encourage the adoption of AI technology in higher education environments. Conclusion: Thus, the implementation of this AI-based attendance system not only improves operational efficiency but also encourages digital transformation in higher education environments. Going forward, this system can be further developed with improved features and scalability for wider adoption across various educational institutions.