The introduction of generative Artificial Intelligence (AI) in education systems has changed the dynamics of the academic processes, particularly for technology-based courses. In this case, an exploratory study of the effects of the use of LLM tools (i.e., OpenAI’s ChatGPT and Google Gemini) on the academic performance of Information Technology (IT) undergraduates at the Faculty of Computing and Information Technology (FCIT), NIIBS Campus. With the help of the scientific approach of empirical research, statistical data were collected through surveys conducted among 49 current undergraduates enrolled in the IT degree program. The research focuses on several important factors related to the penetration of AI tools, their usage frequency and overall effectiveness in terms of GPA achievement. From the study, it was found out that AI technology is widespread among IT students, as 65.3% of participants use both ChatGPT and Google Gemini in multiple platforms. Moreover, 57.1% use them often or very often. From descriptive statistical results, there is evidence of a clear demarcation between task optimization and academic performance. While students exhibit significant levels of operationally positive benefits from speedier completion of assignments (Mean = 3.98/5.00) and enhancement of their quality (Mean = 3.94/5.00), the level of direct impact on terminal GPA performance is notably conservative (Mean = 3.55/5.00). In terms of performance, the study reveals a bell-shaped curve distribution pattern for moderately proficient performance levels in grades (53.0% GPA from 2.5 – 3.0), despite an overwhelming majority perception (89.8%) that AI aids learning positively. There emerges an architectural design of “Academic Integrity Paradox” where efficient structure could induce dependency and ultimately skill degradation. The conclusions drawn by the paper include institutional recommendation for formal inclusion of AI literacy and assessment defense-based approaches.