Evaluate GenAI declaration frameworks empirically and pair them with institutional guidance on which tools are in scope¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The authors propose three strands of future work: empirical evaluation "comparing student declarations against observed workflow data to assess whether declarations track actual behaviour"; pedagogical evaluation of whether the framework changes student reflection, perceived psychological safety around disclosure, and cohort-level patterns of use; and cross-disciplinary adaptation to identify the minimal generalising category set. They also recommend that the declaration "should, thus, be accompanied by clear institutional guidance on which tools are in scope", since IDE autocomplete, writing assistants, and AI-summarising search engines sit on a spectrum with conversational tools.

Design Implications¶

Context¶

Requirements¶

• Institutional guidance defining which tools count as GenAI use must accompany the declaration

Constraints¶

• The framework does not itself detect misconduct, and pairing it with detection tools raises well-documented reliability concerns

Target Learners¶

• Higher education students and educators in Computer Science and, prospectively, other disciplines

Target Learning Goals¶

• Validating that declarations track actual GenAI behaviour
• Improving student reflection and psychological safety around disclosure

Affordances¶

• Domain Specific Genai Declaration Framework

Related Strategies¶

• Replace detector-centred assessment with a four-stage process-oriented model: declared starting point, versioned drafting, process artefacts, and targeted oral verification
• Require students engaging GenAI for feedback to declare the feedback received and how they refined their work in response
• Assess GenAI misconduct evidence against three probative credentials — relevance, credibility and inferential force — before imposing sanctions
• Pursue future research on ChatGPT in higher education to fill the gaps the scoping review identifies
• Increase transparency of AI contributions through disclosure features, tool reminders, and training-data insight

Examples¶

-

Key Sources¶

• Nicholas Micallef, Olga Petrovska. (2026). Structuring Transparency: Developing Domain-Specific Generative AI Declaration Frameworks in Higher Education. International Symposium on Artificial Intelligence and the Transformation of Higher Education Symposium (AITHE). https://www.swansea.ac.uk/staff/nicholas.micallef