AI Tools in Higher Education: Faculty Adoption Grows but Over-Reliance Risks Remain

Faculty at STEM institutions are increasingly turning to AI tools for grading, feedback, and digital assessment — but new research warns that adoption remains uneven and that over-reliance may undermine the critical thinking skills that higher education is designed to build.

A study led by Adel R. Althubyani of Taif University, published in the journal Sustainability, examined how STEM faculty in Saudi Arabia are integrating AI tools in alignment with the United Nations Sustainable Development Goal 4 (quality education). The findings reveal a sector at a crossroads.

Where AI Integration Stands

AI adoption across STEM classrooms scored a mean of 2.71 out of 5 — moderate, and uneven across three key dimensions:

• Assessment is the most developed area. AI tools for automated grading, feedback generation, and digital assessment design are being adopted for their scalability and efficiency.
• Instructional implementation shows moderate uptake. AI-powered platforms, chat tools, and adaptive learning systems are expanding access, especially in post-pandemic environments.
• Instructional planning is the weakest link. Faculty are rarely using AI to design curricula or structure courses — the stage where sustainability principles would have the most durable impact.

The Over-Reliance Concern

Faculty report a high confidence in AI’s potential (mean perception score: 4.00), but significant concern that students are using AI as a shortcut rather than a cognitive tool. Cases of AI-generated submissions — including fabricated references — point to growing academic integrity challenges.

The study also surfaces systemic barriers: poor digital infrastructure, subscription costs for advanced AI platforms, gaps in faculty training, and uneven student digital literacy.

What Needs to Change

Researchers recommend coordinated investment in infrastructure, structured faculty development programs that go beyond technical training to include ethics and pedagogy, and clear governance frameworks addressing data privacy and integrity — all built in partnership between universities, policymakers, and technology providers.

Read the full article on Devdiscourse