Faculty embrace AI tools in STEM classrooms while warning of over-reliance risks

The integration of artificial intelligence into STEM instruction remains uneven and constrained by structural, ethical, and pedagogical challenges, according to new research led by Adel R. Althubyani of Taif University. The study highlights a growing institutional push to align AI adoption with sustainability goals, even as universities struggle to translate technological potential into systemic educational transformation.

The research titled "Responsible AI Integration in STEM Higher Education: Advancing Sustainable Development Goals" and published in Sustainability, examines how STEM faculty in Saudi Arabia are adopting AI tools and how these practices align with the United Nations Sustainable Development Goal 4, which focuses on inclusive and equitable quality education.

AI adoption expands across STEM classrooms but remains moderately embedded

According to the study, AI integration across STEM higher education is progressing, but at a measured and uneven pace. Overall adoption levels remain moderate, with a mean score of 2.71 across key instructional dimensions, including planning, implementation, and assessment.

Among these, assessment practices have emerged as the most developed area. Faculty are increasingly using AI tools for automated grading, generating feedback, and designing digital assessments that improve efficiency and accuracy. These applications are seen as scalable solutions that reduce administrative workload while supporting continuous learning processes.

Instructional implementation, including the use of AI-powered platforms for remote learning and student interaction, also shows moderate uptake. Digital platforms, chat-based tools, and adaptive learning systems are being used to expand access and maintain continuity, particularly in post-pandemic educational environments.

However, instructional planning remains the weakest area of AI integration. Faculty report limited use of AI in designing curricula, structuring lessons, or creating interactive learning environments. This gap is critical, as planning represents the stage where sustainability principles such as inclusivity, resource efficiency, and lifelong learning can be embedded most effectively.

The findings suggest that current AI use is largely operational rather than transformative. While tools are being adopted to improve efficiency, they are not yet systematically reshaping how education is designed or delivered.

Faculty embrace AI potential while raising concerns over learning integrity

Despite moderate integration levels, faculty perceptions of AI are overwhelmingly positive. The study reports a high average perception score of 4.00, indicating strong confidence in AI's ability to enhance teaching quality, improve student engagement, and support modern educational demands.

Educators view AI as a powerful enabler of personalized learning, allowing students to access tailored content, receive immediate feedback, and engage with complex concepts through simulations and adaptive systems. These capabilities are closely aligned with the goals of SDG 4, particularly in promoting lifelong learning and digital skill development.

Faculty express significant concerns about the unintended consequences of AI use. A major issue is the risk of over-reliance, with many educators warning that excessive dependence on AI tools could weaken students' critical thinking, creativity, and problem-solving abilities.

The study highlights a growing tension between efficiency and pedagogy. While AI can streamline processes and improve outcomes, its misuse as a shortcut for information retrieval may undermine deeper learning. Faculty emphasize that AI should function as a support tool rather than a replacement for cognitive engagement.

Ethical concerns also feature prominently. Issues related to academic integrity, data privacy, and the credibility of AI-generated content remain unresolved. Instances of students submitting AI-generated work without verification, including fabricated references, point to the need for stronger oversight and digital literacy.

This dual perspective reflects a mature and cautious approach among educators, who recognize both the transformative potential of AI and the importance of safeguarding educational quality.

Infrastructure, cost, and skills gaps threaten sustainable AI integration

The study identifies a range of systemic barriers that limit the sustainable integration of AI in higher education. Chief among these is inadequate digital infrastructure. Many institutions lack reliable internet connectivity, modern hardware, and access to advanced AI tools, creating uneven learning environments.

Faculty report that weak campus networks and limited technical resources often force both educators and students to rely on personal devices and mobile data. This not only increases financial burden but also exacerbates inequalities, undermining the inclusivity goals of SDG 4.

Cost remains another major obstacle. Advanced AI platforms often require paid subscriptions, making them inaccessible for many institutions. Free versions of tools provide limited functionality, restricting their usefulness in complex STEM applications.

The study also highlights gaps in faculty training. Many educators lack formal preparation in AI integration and rely on self-learning or trial-and-error approaches. This limits their ability to use AI strategically and ethically, particularly in aligning its use with sustainability objectives.

Student readiness presents an additional challenge. Variations in digital literacy mean that some students benefit more from AI tools than others, widening existing disparities. Without structured training, the introduction of AI risks reinforcing inequalities rather than reducing them.

Additionally, concerns about long-term cognitive impact persist. Faculty warn that heavy reliance on AI could reduce opportunities for hands-on learning and independent thinking, which are essential for developing skills needed to address real-world challenges.

AI's role in sustainable education hinges on responsible integration

AI, as the study suggests, is a key driver of sustainable education across environmental, social, economic, and pedagogical dimensions. When it comes to the environment, AI-enabled virtual labs and digital resources reduce the need for physical materials and infrastructure, lowering energy use and waste. Socially, AI has the potential to expand access to education by overcoming geographical and linguistic barriers.

Economically, AI can improve efficiency by automating repetitive tasks and optimizing resource use, allowing institutions to scale education without proportional increases in cost. Pedagogically, AI supports lifelong learning by enabling continuous access to personalized content and feedback.

However, these benefits are not automatic. Achieving sustainable outcomes requires deliberate and responsible integration of AI tools, guided by ethical considerations and aligned with broader educational goals.

Policy and institutional action critical for long-term impact

Meaningful AI integration in STEM education will depend on coordinated action across multiple levels. Here's what the study recommends:

• Investment in infrastructure: Expanding access to high-speed networks, modern devices, and affordable AI tools is essential for ensuring equitable participation.
• Faculty development: Training programs must go beyond technical skills to include ethical awareness, pedagogical strategies, and sustainability principles. Educators need to be equipped not only to use AI but to integrate it in ways that enhance learning outcomes.
• Establishment of governance frameworks that define responsible AI use in education: Clear policies are needed to address issues such as academic integrity, data privacy, and equitable access.
• Collaboration between universities, policymakers, and technology providers: Such partnerships can support the development of context-specific AI solutions that align with educational and societal needs.
• Embedding AI literacy into student curricula: Developing critical thinking and analytical skills in relation to AI is essential for preparing students to navigate an increasingly digital world.