Understanding Generative AI: Implications for Academic Integrity and Citation

Ethical and Productive—Considering Generative Artificial Intelligence Citation Across Learning and Research

Introduction

• Host: Daniel Pfeiffer from Choice and LibTech Insights.
• Speakers:
  • Kari Weaver: Learning, Teaching, and Instructional Design Librarian at the University of Waterloo.
  • Antonio Muñoz Gómez: Digital Scholarship Librarian at the University of Waterloo.
• Context: Discussion on ethical considerations and citation practices for generative AI tools like ChatGPT in academia.

Acknowledgment of Land

• Recognition of the traditional territories where the University of Waterloo is situated.
• Reflection on how citation practices are influenced by colonial approaches to knowledge ownership.

Background of the Project

• Campus Context:
  • Research-intensive university with over 42,000 students.
  • Home to the Waterloo Artificial Intelligence Institute.
• Emergence of Generative AI:
  • Open availability of tools like ChatGPT sparked campus-wide discussions.
  • Initial focus on AI's impact on teaching, learning, and academic integrity.

Focus on Citation Practices

• Purpose of Citation:
  • Creates an information trail and establishes academic connections.
  • Provides standardization and consistency in student assignments.
  • Supports academic integrity through transparency.
• Challenges with AI-generated Content:
  • Difficulty in citing AI-generated outputs.
  • Lack of initial guidance from traditional citation styles.
  • Need for practical solutions for students and faculty.

Ethical Dimensions

• Academic Integrity Concerns:
  • Fear of students using AI to cheat on assignments.
  • Issues with AI detection software misidentifying non-native English speakers.
• Power Dynamics:
  • Discrepancy in the use of AI tools between students and instructors.
  • Data privacy concerns when student work is uploaded to detection software.
• Reproducibility and Accountability:
  • AI outputs are inconsistent; same prompts yield different results.
  • Challenges in preserving AI-generated content for verification.

Citation in Research vs. Learning Contexts

• Research Context:
  • AI tools generally not allowed as authors in publications.
  • AI-generated images discouraged due to reliability concerns.
  • Disclosure of AI use required in methodology sections.
• Learning Context:
  • Adaptation of citation practices to include AI tools.
  • Encouragement for students to be transparent about AI use.

Development of Resources

• Initial Outputs:
  • Created a LibGuide on ChatGPT and generative AI.
  • Developed infographics and annotated prompts illustrating citation practices.
• Ongoing Work:
  • Updating resources to include guidance on citing AI-generated images and videos.
  • Exploring AI tools for literature reviews and knowledge synthesis.
• Campus Collaboration:
  • Formed a campus-wide committee with diverse representation.
  • Contributed to faculty programming and standardized syllabus language.
  • Supported resource development in partnership with other academic units.

Library Initiatives

• Internal Exploration:
  • Monthly sessions on AI tools like Whisper for transcription.
  • Workshops on AI and machine learning in academic libraries.
• Interest Groups and Bibliographies:
  • Formed an interest group on AI within the library.
  • Created a Zotero bibliography with curated readings on AI topics.
• Future Directions:
  • Participation in provincial and federal AI initiatives for academic libraries.

Q&A Session Highlights

• Use of AI in Professional Practice:
  • Librarians using AI tools for brainstorming and instructional design.
• Access to Paywalled Content:
  • AI tools generally cannot access content behind paywalls unless provided by the user.
• Guidance on AI Use in Assignments:
  • Importance of transparency and attribution when students use AI for brainstorming or editing.
  • Encouragement for faculty to discuss AI expectations with students.
• Ethical Considerations:
  • Need to address citation as a colonial practice and explore decolonized approaches.
  • Challenges with integrated AI features in tools and implications for citation.
• Institutional Policies:
  • University of Waterloo currently has no formal policy on AI use.
  • Emphasis on ongoing conversations and collaborative efforts to address AI's impact.

Conclusion

• Recognition of the complexities and rapid development of AI technologies.
• Importance of grappling with ethical, practical, and pedagogical implications.
• Encouragement for open dialogue between faculty, students, and librarians.
• Acknowledgment of the need for adaptable approaches rather than rigid policies.

Note: This summary captures key points from a presentation discussing the ethical considerations and citation practices related to the use of generative AI tools in academic learning and research contexts.