Universities struggle to define AI proficiency as students show uneven adoption across disciplines

Students in engineering and computer science programs are significantly more likely to integrate artificial intelligence into their coursework compared to their peers in humanities and social sciences. Yet regardless of their major, many students report uncertainty about when and how AI use is appropriate in academic settings.

These findings from Virginia Tech research highlight a growing challenge for higher education institutions. If AI competence becomes as fundamental to the modern workplace as spreadsheet proficiency or online research skills, can universities afford to let AI literacy depend on a student’s major, instructor preferences, or personal curiosity?

Three pillars of AI competency emerge from industry insights

The framework for addressing this challenge came from an unexpected source. At CES 2026 in Las Vegas, where 148,000 attendees and over 4,000 companies gathered, Siemens AG CEO Roland Busch outlined three essential components for success in the AI era: technology, domain expertise, and partnerships.

This framework offers universities a roadmap for developing comprehensive AI literacy programs. Technology knowledge represents the obvious starting point – students must understand what AI systems can and cannot do, plus how to use them effectively. They need familiarity with generative AI models, data processing pipelines, and emerging tools.

However, technical knowledge alone proves insufficient for real-world application.

Domain expertise becomes more critical, not less

The second component – domain knowledge – may be even more important than technical skills. AI models are powerful, but they don’t inherently understand the nuances of specific fields like geography, transportation planning, or environmental science.

Research in geospatial data science reveals what experts call “geographic bias” – AI systems often perform better in dense, data-rich urban areas than in rural regions where data is sparse. Without domain expertise, these gaps can go unnoticed and uncorrected.

Key areas where domain knowledge remains essential include:

• Understanding context and limitations specific to each field
• Recognizing when AI outputs may be biased or incomplete
• Knowing how to interpret and validate AI-generated results
• Identifying appropriate use cases within disciplinary frameworks

Partnerships drive interdisciplinary problem-solving

The third component – partnerships – may prove most transformative for higher education. Universities must build productive collaborations across disciplines and beyond campus, including partnerships with communities and industry.

Historically, cross-disciplinary collaboration has been challenging due to different terminology, assumptions, and methodologies in each field. AI tools are beginning to lower these communication barriers by helping translate and clarify unfamiliar technical concepts between disciplines.

This doesn’t replace human collaboration but strengthens it. Meaningful problems are solved by people working together, with AI facilitating rather than substituting for those partnerships.

Access and ethics require institutional attention

Universities face practical challenges in ensuring equitable AI access. Many advanced AI systems require paid subscriptions, and costs can accumulate quickly. While $20 per month may seem manageable for some students, it creates barriers for others.

Institutions are encouraged to expand access to advanced AI infrastructure so that proficiency doesn’t depend on personal financial capacity. The ethical dimension proves equally critical, as empirical research demonstrates that AI outputs contain political, social, and geographic biases.

Students must learn not only how to generate results but how to question them systematically.

Implementation focuses on experiential learning

Universities are well-positioned to lead AI literacy efforts through their existing strengths in experiential learning and collaborative problem-solving. The approach emphasizes project-based work that integrates technical skills with domain challenges and interdisciplinary dialogue.

The question is not whether to bring AI into the classroom – it’s already there. Instead, institutions must focus on preparing students to engage with AI tools meaningfully, combining technology skills with deep domain understanding and strong collaborative partnerships.