Problem

Generative Artificial Intelligence (GenAI) tools are increasingly present in higher education environments, with their adoption largely driven by students. This necessitates that instructors adapt to technologies already integrated into classroom practices. While some faculty have embraced GenAI for various pedagogical purposes, such as content creation and assessment support, others approach these tools with significant caution. Their concerns primarily revolve around potential negative impacts on student learning, the validity of traditional assessment methods, and upholding academic integrity. Therefore, a comprehensive understanding of faculty perspectives is crucial for developing effective pedagogical strategies and robust institutional policies to guide the responsible integration of GenAI.

Method

To investigate STEM faculty perspectives on GenAI in teaching and learning, a qualitative study was conducted utilizing focus group discussions. Twenty-nine STEM faculty members from the College of Science and Engineering at San Francisco State University participated. The study involved seven remote focus group sessions, each lasting approximately 90 minutes with three to five participants. The discussions were structured around three core research questions: how faculty integrate GenAI into course design, the observed benefits and challenges for student learning, and the institutional support required for effective GenAI adoption. Data analysis involved anonymizing Zoom transcripts and AI-generated summaries, which were then cross-referenced and manually verified by the research team.

Results

The study revealed that 93% of participating faculty actively use GenAI tools for teaching and learning. Their applications fall into three main categories: facilitating course design and preparation (e.g., generating quizzes, developing rubrics), directly supporting student learning (e.g., brainstorming, coding assistance, designing critical evaluation assignments), and assisting with administrative tasks (e.g., summarizing emails). Faculty observed that GenAI can help bridge technical gaps for students, accelerate project development, and act as a "private teaching assistant" by providing immediate responses. However, significant challenges were also identified. These include students facing an increased burden in crafting effective prompts, the uneven quality of AI-generated content requiring careful review, and a concern that over-reliance on GenAI might bypass critical thinking and problem-solving development. Academic integrity and plagiarism were major concerns, with faculty noting the difficulty of detecting AI-generated work and the unreliability of current detection tools. As a result, some faculty are reverting to traditional, controlled assessment methods like in-class exams or designing assignments that explicitly require critical engagement with AI outputs.

Implications

The effective integration of GenAI in higher education demands a fundamental rethinking of assessment practices, pedagogical approaches, and institutional governance. Key implications include the need for extensive professional development workshops for faculty, covering GenAI fundamentals and prompt engineering, alongside the creation of centralized resource repositories. Dedicated AI consultation services and communities of practice are also essential. Institutions must develop clear, consistent policies that balance university-wide guidelines with departmental autonomy, addressing academic integrity, intellectual property, citation standards, and ethical considerations such as accessibility and bias. The findings underscore that while GenAI offers opportunities for educational enhancement, its responsible adoption requires thoughtful planning, sustained investment in faculty development, and a cautious approach to implementing permanent institutional changes, given the rapidly evolving nature of the technology and its potential impact on future workforce skills.