This integrative literature review examines how undergraduate and graduate students in STEM disciplines utilize generative artificial intelligence (GAI) tools, such as ChatGPT, Gemini, and Copilot, to support learning. Synthesizing prior research, the review employs an AI-augmented methodology that combines traditional database searches with human-verified AI-assisted support. The review analyzes the (1) theoretical perspectives and frameworks; and (2) reported learning outcomes guided by the multidimensional Universal Competency Taxonomy (UCT), including STEM, learning, practice, value, meta, community of practice and people competencies. Findings reveal that technology acceptance models (TAM, UTAUT) dominate the theoretical landscape, while constructivist, self-regulated learning, and cognitive load theories offer complementary perspectives on AI’s pedagogical role. Empirical evidence indicates generally positive effects on academic performance, motivation, and engagement when AI integration is intentionally structured, with frameworks like the Guidance-based ChatGPT-assisted Learning Aid (GCLA) yielding substantial learning gains. However, unstructured or passive reliance on AI may undermine critical thinking and independent problem-solving, highlighting the need for explicit scaffolding and metacognitive instruction. The review identifies equity concerns related to AI literacy and access, emphasizing that targeted training and infrastructure are essential to prevent widening achievement gaps. This paper argues that the key challenge for STEM educators is not whether to permit generative AI, but how to design learning environments that foster productive, ethical, and reflective engagement with these tools. The review calls for future research to expand theoretical frameworks, address understudied competency dimensions, and examine long-term impacts of AI use on professional formation, particularly in diverse and resource-constrained contexts.
http://orcid.org/0000-0002-1907-9463
Florida International University
[biography]
http://orcid.org/https://0000-0002-7821-9059
The Pennsylvania State University
[biography]
The full paper will be available to logged in and registered conference attendees once the conference starts on June 21, 2026, and to all visitors after the conference ends on June 24, 2026