By Eddie Lin and Roshan Bharwaney
The gap between higher education and workplace demands has been evident for years (Hansen, 2021; Weissman, 2024; Stephenson & Swift, 2025). Indeed, some higher education leaders argue against employability as the core purpose or value of higher education and prefer to focus on developmental rather than vocational goals (Tight, 2023). Declining interest in university education (Fry et. al., 2024; Anderson, 2023) reflects growing concerns about the return on investment of traditional degrees, particularly in terms of employability. The rise of artificial intelligence (AI) and automation deepens this challenge: how can higher education prepare graduates for a labor market where AI fluency is as critical as subject-matter expertise? And, equally, how must employers adapt their expectations and support systems to this new reality?
To answer these questions, we must first examine how AI’s roles tend to differ in higher education and the workplace.
AI’s Roles in Higher Education & in the Workplace
In universities, AI use by students is often discouraged or tightly restricted under the banner of academic integrity. For some courses, with an instructor’s permission, students have to disclose how they have used AI for coursework, which leads to new challenges with assessment and evaluation. In contrast, employers increasingly expect frequent AI use, viewing it as a lever to boost productivity, reduce costs, and improve decision-making, among other uses.
The traditional higher education model has long emphasized content mastery: Students acquire disciplinary knowledge and demonstrate competence through methods such as essays, exams, and research. The focus is on cultivating domain expertise and intellectual rigor before entering the workforce. Yet, employability often hinges on practical abilities such as communication, collaboration, applied problem-solving, and program management. These skills are difficult to measure or aren’t often prioritized with current academic assessments and often fall outside disciplinary silos, leaving graduates under-prepared for workplace realities.
Dr. Shu-Yi Hsu, a senior instructional designer at Columbia University, describes attitudes to AI in higher education and a recommended approach: “Attitudes toward AI in higher education range from strong enthusiasm to deep skepticism. Faculty worry about fairness in grading and whether reliance on AI erodes genuine learning and critical thinking. AI should augment, not replace human teaching. Students must have AI literacy and [know] how to use AI responsibly while recognizing its limitations.”
Historically, this mismatch could be bridged by employers through methods such as structured onboarding, apprenticeships, internships, and on-the-job learning with gradual exposure to more complex tasks. AI disrupts this bridge in two ways. First, AI performs much of the entry-level work that once gave junior employees opportunities to build practical skills and confidence while deepening their domain expertise, and second, the new highest human value in AI-enabled workplaces lies in cognitive skills to supervise AI: asking the right questions, evaluating outputs, and refining solutions (Bharwaney & Lin, 2025). Yet, only employees with strong domain knowledge and professional experience are positioned to exercise these high-level skills.
This dynamic has produced two conundrums: Many organizations are reducing entry-level hiring because AI has shown potential to execute similar tasks efficiently (Temkin, 2025), while simultaneously struggling to motivate and guide existing employees to adopt AI and apply their expertise in new ways (Law, 2025)
Silver Lining: The Opportunity for Higher Education
Inspirational professors, structured curricula, and formal assessment of academic achievement remain higher education’s powerful assets. The opportunity lies in harmonizing these academic strengths with the workforce’s evolving demand for AI-enabled skills.
For example, in both university and the workplace, AI can be used for summarizing, drafting, and accelerating research, with shared expectations for ethical use, validation, and critical judgment. While higher education emphasizes academic rigor, declarative knowledge, critical thinking, and intellect, and workplaces prize speed, influence, and execution, the skills of swiftly grasping and applying new knowledge and pursuit of original ideas converge.
Dr. Yoo Kyung Chang, clinical professor & academic director of information systems and technology at NYU School of Professional Studies, said, “AI should be treated as a cognitive tool in higher education so learners can focus on higher-level thinking, much like how calculators support advanced mathematics. Students must master intelligent tasks themselves before offloading them to AI, practicing critical thinking to decide what to do, what to delegate, and how to collaborate with these tools.”
Higher Ed Curricula Must Evolve
AI’s potential in higher education is to make training of students’ structured skills (i.e., those that follow rules and are explicitly taught) more efficient and effective, such as memorization, understanding, and knowledge application in academic or professional domains. With that in mind, students and faculty can focus more on students’ unstructured skills, which are those that are emergent, adaptive, and generally developed through experience. Examples are problem solving, critical thinking, questioning assumptions, innovation, adaptability, ideation, integrating knowledge across disciplines, handling ambiguity, collaboration, leadership, and designing independent research.
For higher education to remain relevant, curricula must evolve. Here are some overarching recommendations for directions in higher education to bridge the skills gaps between universities and workplaces:
- AI ethics and safety: Prepare students to navigate issues of fairness, bias, privacy, and societal impact.
- Tackling complex questions: Emphasize open-ended challenges that blend structured and unstructured skills and reduce reliance on standardized tests and repetitive drills.
- Critical thinking: Develop new assessments for judgment, creativity, and metacognition—essential to supervise AI outputs.
- Human-AI synergy: Embed AI fluency across all disciplines, encouraging students to find the niches where human value is maximized.
- Industry connection: Maintain close industry partnerships and collaborations including open innovation opportunities and collective intelligence approaches (Bharwaney & Sleeva, 2024).
Experiential learning and communities of practice are central to this vision. Internships, simulations, and cross-disciplinary projects can help students practice human-AI collaboration, resilience, and decision-making in environments that mirror the workplace’s ambiguity and complexity.
Conclusion: Shaping AI Use
Universities that condemn the use of AI by students risk isolating themselves from the realities of today’s workplace, where interns and new hires are expected to be or quickly become adept at using AI for routine tasks and complex projects. By thoughtfully integrating AI use into curricula and coursework, emphasizing both technical proficiency and principles of responsible and ethical use, higher education can evolve its teaching and assessment methods to ensure that the workforce of tomorrow retains its distinctly human judgement, creativity, and values in an era increasingly shaped by technology. Academia is uniquely poised to help ensure that technological adoption enhances, not erodes, the human dimensions of work.
References
- Anderson, J. (2023). How America Started to Fall Out of Love With College Degrees, Time.
- Bharwaney, R., & Lin, E. (2025). Mind the Skills Gap: Moving from AI Myths to AI Supervision Skills. Cambridge Journal of Artificial Intelligence.
- Bharwaney, R., & Sleeva, S. (2024). Navigating New Frontiers for AI Talent Access: Open Innovation and Collective Intelligence. Chief Talent Officer.
- Fry, F., Braga, D., & Parker, K. (2024). Is College Worth It? Pew Research Center.
- Hansen, M. (2021). The U.S. Education System Isn’t Giving Students What Employers Need. Harvard Business Review.
- Law, M. (2025). Kyndryl: 71% of Workforces Unprepared for AI Deployment. Technology Magazine.
- Stephenson, D., & Swift, J. (2025). The Bridgeable Divide Between Higher Education and Work. Governing.
- Temkin, M. (2025). AI May Already Be Shrinking Entry-Level Jobs in Tech, New Research Suggests. TechCrunch.
- Tight, M. (2023). Employability: A Core Role of Higher Education? Research in Post-Compulsory Education, 28(4), 551–571.
- Weissman, S. (2024). A ‘Great Misalignment’ Between Credentials and Jobs. Inside Higher Ed.
Image credit: Feodora Chiosea
