UPSA Redefines Programming Pedagogy to Counter AI-Generated Code

The rapid proliferation of generative artificial intelligence has created an existential inflection point for computing education. As Large Language Models (LLMs) become capable of producing syntactically correct code in seconds, the traditional metrics of student success—executable output and algorithmic accuracy—are becoming obsolete. In response, the University of Professional Studies, Accra (UPSA) has implemented a pioneering instructional model designed to pivot the focus of programming education toward higher-order cognitive skills: reasoning, traceability, and verification. This shift acknowledges that in an AI-saturated market, the value of a software engineer lies not in writing code, but in the ability to conceptualize, defend, and refine complex systems.

During the first semester of the 2025/2026 academic year, UPSA introduced this project-centered architecture to its Level 300 BSc Information Technology students. Led by Dr. Augustina Dede Agor, the initiative targets students in the IT4 and IT5 cohorts, moving away from rote memorization and toward a 'disciplined programming competence.' The model is built on the premise that while AI can provide the 'what' of a program, the human developer must remain the master of the 'why.' This necessitates a curriculum that prioritizes the documentation of logic and the justification of design decisions over the mere delivery of a functional script.

“

The rapid proliferation of generative artificial intelligence has created an existential inflection point for computing education.

Central to the UPSA model is a suite of reinforcing mechanisms that mirror professional software engineering environments. These include continuous assessment embedded directly into lectures, instant evaluation of task execution, and the use of repository workflows. Perhaps most critically, the model introduces 'live technical defense' of implementation logic. By requiring students to verbally and technically defend their code, the university ensures that students are not merely prompting an AI to solve their assignments, but are instead deeply engaged with the underlying architecture of their work. This creates a culture of technical accountability that is often missing in traditional, output-oriented grading systems.

This pedagogical evolution is deeply rooted in the constructionist theories of Seymour Papert. Papert famously argued that effective computing education must create environments where learners actively construct knowledge rather than passively receiving it. By centering the course on iterative mini-projects and incentive-based performance reinforcement, UPSA is operationalizing Papert’s vision for the modern era. The focus on 'incentive-based reinforcement' tied to quality thresholds suggests a move toward competency-based education, where students are rewarded for the rigor of their engineering process rather than just reaching a binary 'pass/fail' state on code execution.