The integration of AI-based advisory systems in higher education has transformed student advising by improving accessibility, efficiency, and accuracy. This study evaluates a GPT-based robo-advisory system designed to assist students with academic enquiries, administrative processes, and career advice. The system was trained on 29 module handbooks, general rules, and FAQs to ensure institution-specific accuracy. A test of 100 sample queries across different categories demonstrated 100% overall accuracy, reinforcing its reliability in providing correct and contextually relevant answers. Compared to traditional chatbots, the system excelled at interpreting complex queries, providing detailed academic insights, and streamlining university processes. In line with existing literature, the findings confirm that AI-driven advising improves student engagement and administrative efficiency, complementing previous research on chatbots in education. However, challenges remain, including the integration of predictive analytics, multilingual support, and ethical considerations, particularly when dealing with sensitive student issues. While the model successfully automates academic advising, it does not replace human advisors in cases that require emotional intelligence and ethical judgement. Addressing biases in AI-generated responses through refined training data and improving computational efficiency with scalable processing strategies are crucial for ensuring fairness and widespread implementation. Future developments should focus on hybrid models that combine AI-driven insights with human expertise to optimise student support. This study contributes to ongoing discussions about AI in higher education and advocates for adaptive, data-driven, and ethically responsible student advising systems.