Skip to content

References: Responsible and Ethical Use of AI

  1. Ethics of Artificial Intelligence - Wikipedia - Broad survey of AI ethics covering bias, privacy, accountability, transparency, environmental impact, and regulation. Gives students a vocabulary and map of the risks this chapter turns into practical lab rules.

  2. Academic Integrity - Wikipedia - Defines academic integrity, plagiarism, and misconduct, with sections on institutional policy and generative AI. Useful background for the chapter's rules on honest representation, citation, and disclosure of machine-assisted work.

  3. Algorithmic Bias - Wikipedia - Explains how systematic errors arise from data, design, and deployment, with documented historical examples and regulatory responses. Supports the chapter's checks for biased literature surveys and uneven performance across groups.

  4. Weapons of Math Destruction (First Edition) - Cathy O'Neil - Crown - Popularized three tests for harmful algorithmic systems: opacity, scale, and damage. These give researchers a concrete checklist for judging when an AI-assisted process needs scrutiny, accountability, and transparency.

  5. Fairness and Machine Learning: Limitations and Opportunities (First Edition) - Solon Barocas, Moritz Hardt, and Arvind Narayanan - MIT Press - Organizes fairness criteria into independence, separation, and sufficiency and shows why they conflict. This formal framing clarifies the chapter's bias and fairness concerns when AI is applied to data about people.

  6. AI Risk Management Framework - NIST - U.S. framework for managing AI risks and building trustworthiness, with a playbook and generative AI profile. A model for turning principles such as accountability and transparency into the structured lab AI policy drafted here.

  7. Energy and Policy Considerations for Deep Learning in NLP - arXiv (Strubell, Ganesh, McCallum) - Estimates the energy and carbon cost of training large language models and proposes mitigations. Grounds the chapter's environmental cost discussion with measured figures and recommendations for proportionate model use.

  8. ICMJE Recommendations - International Committee of Medical Journal Editors - Widely adopted publishing standards, including defining authorship roles and a section on AI use by authors, reviewers, and editors. Shows how venues justify rules that AI cannot be an author and that use is disclosed.

  9. Living Guidelines on the Responsible Use of Generative AI in Research - European Commission - Guidance for researchers, research organizations, and funders on using generative AI while protecting research integrity. A worked example of institutional policy covering disclosure, confidentiality, and verification.

  10. Recommendation on the Ethics of Artificial Intelligence - UNESCO - First global standard on AI ethics, adopted by 193 member states, setting out values, human-rights-based principles, and policy areas. Provides an international reference for privacy, fairness, human oversight, and environmental responsibility.