Assisted but Unguided: AI Tool Integration, Academic Performance, And Institutional Policy Gaps Among Filipino College Students

Abstract This research investigated how Artificial Intelligence (AI) tools can affect the performance of college students in one of the state universities in Pangasinan, Philippines. The research design was a convergent parallel mixed-method survey design, in which 300 first- year to fourth-year college students were surveyed in five focus group discussions of five academic colleges to produce both quantitative and qualitative data. A survey instrument developed by the researcher was used to collect quantitative data and semi-structured focus group discussions (FGDs) were used to collect qualitative data. Findings indicated that 91.3 percent of respondents adopted ChatGPT as the major tool of AI. The results also indicated the statistically significant difference in the increase of the General Weighted Average (GWA) of students before (M = 83.72) and after (M = 87.03) regular use of AI tools (p =.001). Generally, students were positive (M = 4.10) that AI tools had positive effects on their academic performance, especially in comprehension of lesson material, quality of writing and research effectiveness. Nevertheless, qualitative results raised an issue related to over-reliance, academic integrity, and the lack of a formal institutional policy of AI. The article concludes that although it is evident that AI tools can enhance academic performance outcomes, their uncontrolled application has serious pedagogical consequences. The research proposes the creation of AI literacy curriculum, institutional policies about AI usage, and faculty training programs according to the existing and future educational technology standards.