Responsible Use of AI in Academic Research

Team Name

Cogent

Use of AI Tools

Which tool(s) did you use?

• ChatGPT

• Google Gemini

For what purpose(s)?

• Research support

• Structuring ideas

• Refining arguments

• Drafting and editing content

• Improving clarity and organization

• Reviewing grammar and language flow

How did these tools support your work?

The AI tools helped us organize our ideas more effectively, improve the structure and readability of the document, and refine our arguments. They also assisted in brainstorming implementation strategies and identifying gaps in our analysis. However, all factual claims, references, and final decisions were manually reviewed and verified by the team.

External Feedback & Contributions

• Dr. Paras Chawla – Supervisor – Mentoring, guidance, and feedback throughout the project

• Dr. Christelle Scharff – Expert Reviewer – Feedback on implementation clarity and framework practicality

• Nicolas Lepotier – Reviewer – Feedback regarding student compliance and incentive structures

• Milagros Espejo Bocanegra – Peer Reviewer – Discussion and feedback on AI trust, verification behavior, and learning processes

Initial Contribution

https://drive.google.com/file/d/1MkNfQVcCgY--elN2jwRwKE0u-CIvumMz/view

Final Contribution

https://drive.google.com/file/d/1TDrFLP6f7RB412YxvLE5JEp0vzmv1gPz/view

Our final contribution focuses on the responsible use of generative AI in academic research through the development of the VERIFIED AI RESEARCH PROTOCOL (VARP). The framework promotes transparency, verification, and responsible AI usage rather than outright restriction.

VARP includes:

• AI usage logging

• Mandatory citation verification

• Tiered AI disclosure levels

The framework is designed to be low-cost, scalable, and practical for university implementation.

Reflection on the Process

During Phase 2, our contribution evolved significantly through discussions, expert feedback, peer comments, and keynote insights.

Initially, our work focused mainly on the issue of AI hallucinations and fabricated citations. However, feedback from reviewers encouraged us to expand the framework beyond technical risks and address broader dimensions such as academic integrity, equity, institutional policy gaps, environmental concerns, and student learning behavior.

Dr. Christelle Scharff’s feedback helped us strengthen the implementation section by making the workflow more practical and institution-ready. Nicolas Lepotier’s comments pushed us to think critically about student honesty and incentive structures within disclosure systems. Milagros Espejo Bocanegra’s discussion influenced our reflection on how students decide whether to trust AI outputs and how verification habits develop over time.

As a result, we added:

• A phased pilot rollout strategy

• Faculty training mechanisms

• Trust and surveillance safeguards

• Failure scenarios and mitigation strategies

• More detailed implementation conditions

• Stronger reflection on equity and cognitive impacts

This process made our proposal more realistic, balanced, and evidence-driven. It strengthened our understanding that responsible AI governance in education cannot rely on simple bans or unrestricted usage, but instead requires transparent, testable, and adaptable institutional frameworks.

Video Pitch

https://drive.google.com/file/d/1h4Tbg2sPw42sCvIVi9IMX9KQIy4-41Fm/view