Beyond Detection: A Proposed Framework for Redefining Academic Integrity in the Generative AI Era

Resumen

The rise of generative AI has created a crisis of originality in higher education, rendering traditional definitions of cheating and plagiarism nearly obsolete. AI introduces new ways of thinking that challenge the foundations of academic work and knowledge creation. University responses have been inadequate, swinging between unenforceable bans and flawed detection tools that raise ethical concerns, including false positives and biases against non-native English speakers. Such approaches fail to engage with the new reality and instead deepen distrust, ultimately undermining both effective teaching and academic integrity.

This ongoing project argues that universities need a novel, comprehensive, and forward-looking strategy to navigate this tumultuous environment ethically. It proposes a new three-pillar framework for academic integrity, designed specifically for higher education, that replaces the purely punitive model with one grounded in intellectual honesty and responsible collaboration with AI.

The proposed framework rests on three interconnected pillars:

1. Policy Reform: This pillar focuses on rewriting academic integrity policies to address AI explicitly. It shifts the emphasis from "plagiarism" (copying text that already exists) to "unauthorized use" and requires a tiered system of AI use (AI-Free, AI-Assisted, AI-Centric) with clear, required rules for citing and disclosing AI use in all assignments.
2. Assessment Redesign: This significant change in teaching requires rethinking how assessments are structured so they are less susceptible to AI interference. Key strategies include grading research proposals and drafts based on the process rather than the product, giving more high-stakes tests in class (oral defenses and presentations), and focusing on assignments that require experiential or personalized analysis, as those are skills that only humans have.
3. AI Literacy: This pillar calls for mandatory education for all university stakeholders, recognizing that policy alone is ineffective without understanding. It urges institutions to promote faculty training and professional development focused on course redesign and to establish required student workshops that address the ethical capabilities, limitations, and responsible use of generative AI as a collaborative instrument.

This study compares global university policies and conducts qualitative research to refine a practical framework that helps institutions build a culture of responsible AI aligned with higher education’s humanistic principles.