Key points are not available for this paper at this time.
ABSTRACT The rapid integration of artificial intelligence into higher education has generated both enthusiasm and concern among STEM educators. Current discourse has largely emphasized risks related to academic integrity, student overreliance, and assessment validity. While these concerns are justified, an exclusive focus on restriction risks overlooking the potential of artificial intelligence as a constructive pedagogical implement, particularly in resource-limited academic environments. In this perspective, we argue that generative artificial intelligence, when integrated transparently and ethically, can function as an equity-enhancing tool in STEM education by extending mentorship, supporting scientific communication, and facilitating active learning. Drawing on graduate-level teaching and laboratory mentoring experiences in life sciences disciplines within a Global South context, we illustrate how artificial intelligence can help address structural constraints such as limited faculty time, language barriers, and uneven access to academic support. We also examine challenges associated with artificial intelligence adoption, including dependency, unequal access, and institutional resistance, and outline practical strategies for responsible integration. Rather than framing artificial intelligence as a threat to learning, we propose that STEM educators view it as an opportunity to modernize pedagogy, foster artificial intelligence literacy, and better prepare students for contemporary biological research.
Espinosa-Hernández et al. (Fri,) studied this question.