Abstract
The ELEVATE (Enhancing Learning Experiences Via AI Techniques) program responds to the rapid, unstructured adoption of generative artificial intelligence (AI) in higher education by developing and evaluating personalized AI tutors that function as instructional scaffolds rather than shortcut solutions. As student use of AI continues to outpace curricular guidance—particularly at Minority Serving Institutions (MSIs)—ELEVATE addresses the need for evidence-based strategies that preserve critical thinking, reduce misinformation, and promote equitable learning outcomes. Implemented at California State University, Bakersfield (CSUB), a federally designated Hispanic Serving Institution (HSI), the program is grounded in constructivist learning theory, Vygotsky’s Zone of Proximal Development, and computational thinking (CT). Conversational AI is positioned as a “Knowledgeable Other” that supports dialogic, step-by-step learning through guided problem decomposition and reflection, rather than holistic answer generation. Over an 18-month period, faculty across STEM and non-STEM disciplines will redesign courses to integrate AI tutors aligned with learning outcomes, providing more than 200 students—many of whom are low-income, first-generation, or high-workload—with hands-on experience using industry-relevant AI tools. A mixed-methods evaluation will examine learning outcomes, student perceptions, and equity impacts, with particular attention to whether structured AI tutoring mitigates or exacerbates the digital divide for minority students. Ethical considerations, including algorithmic bias, accessibility, and data privacy, are embedded throughout the program. By pairing AI literacy with intentional instructional design and faculty development, ELEVATE aims to shift perceptions of AI from a threat to academic integrity toward a validated pedagogical tool, producing a transferable model for equitable and responsible AI integration in postsecondary education.