> Research Guide - Generative AI Literacy

GenAI spans a broad spectrum. While LLMs (like ChatGPT) and Diffusion Models are the most prominent, GenAI includes diverse technologies. Model selection depends on data type (text/image/audio), output quality, and efficiency needs.
 

Introduced by the Teaching and Learning Development Office (TLDO), HKSYU staff and students can:

• Use HKSYU ChatGPT to facilitate teaching, learning, and assessment tasks.
• View HKSYU ChatGPT Tutorials to learn more about HKSYU ChatGPT and the University's guidelines.
• Understand the Guidelines on the Use of Generative AI in Teaching, Learning, and Assessment (English Version | Chinese Version) to make appropriate adaptations to incorporate the use of AI in assignments.

• Learning Resources | SYULC English Trainer
• Workshop Videos

                                                                                                                                                                                                                   
 

Evaluating Information: Applying the CRAAP Test by Meriam Library, California State University, Chico

The CRAAP test is a useful framework for evaluating the reliability and credibility of sources used in research.

Credible sources are essential in ensuring the accuracy of information, strengthening arguments, and reinforcing conclusions. With the vast amount of available information, distinguishing trustworthy sources requires careful scrutiny—a vital skill in information literacy.

To assist researchers in this effort, librarians at California State University introduced the CRAAP Test in 2004. This method provides a structured approach to assessing a source’s Currency, Relevance, Authority, Accuracy, and Purpose, helping individuals make informed decisions about the credibility of information. It can also be applied by end-users to evaluate sources generated by AI.

Using the CRAAP test to evaluate AI-generated content ensures that the information you rely on is trustworthy, relevant, and accurate. It’s always a good practice to cross-check AI-generated responses with authoritative sources before using them for research or decision-making.

You should cite generative AI (such as HKSYU ChatGPT) whenever you use its output in your work, particularly for direct quotes, paraphrased ideas, or data it provides. Proper attribution maintains academic integrity and helps avoid plagiarism.

Always follow your instructor's or publisher's citation guidelines.

Additionally, verify AI-generated facts, as they may be inaccurate due to hallucinations.

Besides providing a guide to citation styles and citation management tools, we also gather the following methods to cite AI-generated materials in APA and Chicago styles. Here is a breakdown of the current recommendations. Keep in mind that these guidelines are evolving, so it is always a good idea to check the latest editions of the style manuals or official style blogs for updates.

If you want to know more about examples of citation formats for generative AI, you can also refer to Appendix C: Citation Formats of Generative AI in APA and MLA, in the Hong Kong Shue Yan University Guidelines on the Use of Generative AI in Teaching, Learning, and Assessment by the TLDO. 

 

Generative AI is playing a growing role in scholarly publishing, presenting both opportunities and challenges. Publishers have adopted varying policies and practices to address AI-related concerns. When submitting manuscripts, authors should ensure compliance with publisher guidelines and carefully consider the permissions they grant for the reuse of their work in AI applications.

Elsevier > Generative AI policies for journals

Generative AI Policies for Journals

Elsevier has outlined clear policies on the use of generative AI and AI-assisted technologies in scientific writing and publishing. Key policies include:

• AI tools can only be used to improve readability and language but not to replace key authoring tasks like generating scientific insights or conclusions.
• Authors must disclose their use of AI tools in their manuscripts, and a statement will appear in the published work to ensure transparency.
• AI tools cannot be credited as authors or co-authors, as authorship requires accountability and responsibilities that only humans can fulfill.
• AI-generated or AI-altered images are not permitted in submitted manuscripts, except when AI is part of the research design or methods.
• This guidance clarifies the responsibilities of reviewers and authors, with a detailed FAQ provided for reference.

Emerald > Emerald Publishing’s stance on AI tools in content creation and the peer review process > Emerald Publishing’s stance on AI tools and authorship

Emerald Publishing’s Stance on AI Tools and Authorship

Emerald Publishing acknowledges the growing role of AI in research but emphasizes transparency, accountability, and ethical use. Key announcements include:

• AI tools cannot be credited as authors since they lack accountability for the work generated.

• Authors must disclose their usage of AI tools during submission.

• AI tools can assist in copy-editing to improve language and readability.

• AI tools are prohibited from creating original content or being involved in the peer review process.

• AI-generated images are not accepted for publication.

• Emerald Publishing emphasizes transparency, integrity, and high-quality academic content while adapting policies as AI technology evolves.

Oxford Academic > Author use of Artificial Intelligence (AI)

Author Use of Artificial Intelligence (AI)

Oxford University Press (OUP) has issued a policy on AI use in academic book publishing, outlining guidelines for authors.

These guidelines may evolve as AI technology advances, so authors should check for updates during submission.

• AI cannot be listed as an author or perform core writing tasks like generating arguments or conclusions. Any AI-generated content requires OUP's prior written permission and may need replacement with human-created content.
• Authors are fully responsible for all content accuracy, originality, and ethics - including AI-assisted portions - and must verify proper attribution while avoiding bias or conflicts.
• All AI use in content creation must be disclosed to the Acquisitions Editor and properly cited according to style guidelines.
• Scholars advising OUP must submit only vetted, expert-approved material and balance AI-generated ideas with their professional judgment.
• Manuscripts or project materials must never be uploaded to generative AI tools to protect intellectual property.

Sage Journals > Using AI in peer review and publishing

Sage's Guidelines on AI in Peer Review and Publishing

Authors must disclose any use of AI tools that generate content (such as text, images, or references). Original sources should always be cited instead of AI tools as primary references. AI-assisted or AI-generated submissions must be disclosed at the time of submission for editorial evaluation. Authors are required to adhere to Sage guidelines, specifically to:

• Authors must clearly state in the methods or acknowledgments section which language model was used and its specific purpose in the research.
• All AI-generated content and citations must be carefully checked for accuracy, validity, and appropriateness, with any errors or biases corrected.
• Authors should verify that the AI hasn't plagiarized by checking original sources for any reproduced text.
• Every factual claim and citation must be confirmed as authentic, as AI may generate false information or nonexistent sources.
• AI tools like ChatGPT cannot be listed as authors on any submission.
• Authors maintain full responsibility for all aspects of their manuscript, including any AI-assisted content.

Taylor & Francis > AI Policy

Taylor & Francis AI Policy

Authors are fully accountable for ensuring the originality, validity, and integrity of their submissions. All content must meet rigorous scholarly standards and represent the author's own intellectual work, with proper validation of research and analysis.

Taylor & Francis supports responsible use of generative AI tools for specific applications, provided they adhere to strict standards of data security, confidentiality and copyright protection. Approved uses include:

• idea generation and exploration
• language improvement
• enhanced literature searches using LLM-powered tools
• research classification
• coding assistance

All AI-generated content must be carefully reviewed for accuracy. Journal authors must follow editorial policies and publishing ethics, while book authors must comply with specific publishing guidelines. Importantly, AI tools cannot be listed as authors under any circumstances.

Any use of generative AI assistance must be explicitly acknowledged within the manuscript or book. Authors must transparently disclose how AI tools were employed in their research or writing process.

Wiley > Artificial Intelligence General Guidance & Best Practice

Artificial Intelligence General Guidance & Best Practice

Wiley allows AI in manuscripts if authors:

• uphold standards
• protect rights
• disclose use

Full accountability remains with authors.

Any use of GenAI tools in manuscript development must be fully disclosed in either the Methods section, a dedicated disclosure statement, or Acknowledgements, with specific details about how the tool was used.

Authors bear complete responsibility for verifying the accuracy of all AI-generated content and must properly cite any sources the AI tool references.

GenAI tools are strictly prohibited from being used to create, alter, or manipulate original research data and results.

Basic editing tools for spelling, grammar, and general language improvements are exempt from these disclosure requirements.

Journal editors retain final authority to determine whether any GenAI use in a submission or published article complies with the publication's editorial policies.

In peer review contexts, GenAI tools may only be used minimally and with appropriate safeguards to maintain review integrity.

 

   

• Exercise caution when using content that may involve others' copyrighted works.
• Prioritize original creations whenever possible.
• Comply with the terms of service/use of the AI system or website.
• Verify whether the AI developer has obtained proper licensing from copyright owners to use protected works for training their generative tools.
• Understand Creative Commons(CC) Licenses and Generative AI. Utilize CC works and adhere strictly to their licensing terms.
  • Sources for AI-Generated Works under CC Licenses
    • Wikimedia Commons – A free media repository that allows users to share and access a vast collection of images, audio, and video files under open licenses (e.g., CC License, GNU Free Documentation License, Artistic License, and Free Art License) or in the public domain.
      • Search "AI-generated Creative Commons" on Wikimedia Commons.
    • Openverse – A search engine designed to help users discover free stock photos, images, and audio available under Creative Commons licenses or in the public domain.
      • Search for a keyword and use filters to sort by CC license type.
    • DeviantArt – An online art community where artists share and sell diverse creative works, including digital art, photography, and AI-generated content, under various licenses (though Creative Commons options are limited).
      • Search for AI art, then verify CC licenses individually in the "Details" tab.
    • Stable Diffusion Gallery  – A collection of AI-generated images created using the Stable Diffusion model, where users can explore, share, and sometimes download artworks.
      • Be primarily licensed under CreativeML OpenRAIL++. Some AI-generated works may be released under CC licenses, depending on the creator or platform. For example, projects like CommonCanvas train diffusion models using images licensed under Creative Commons.
    • MidJourney Gallery  – A collection of AI-generated images created through text prompts, the MidJourney Gallery showcases diverse artworks from its community, though most content retains full copyright protection unless individually marked with Creative Commons licenses by creators.
      • Search MidJourney's gallery manually for license declarations in descriptions (despite lacking official CC tagging tools), as some creators may self-report alternative licenses while most default to copyright.
    • OpenGameArt – A free and open-source media repository that provides game developers with assets, including sprites, textures, sound effects, and music, all available under permissive licenses for creative projects.
      • Search for AI-generated assets using CC license filters.

• Key Considerations
  • ✔ Verify the License – CC0 = public domain (no attribution), CC BY = requires credit, CC BY-NC = non-commercial use.
  • ✔ Check Platform Rules – Some sites ban AI-generated content, while others embrace CC licensing.
  • ✔ Ethical Use – Ensure the AI tool’s terms allow CC licensing (e.g., Stable Diffusion permits CC0; DALL·E 3 has restrictions).

Loading ...

Loading ...

Apple Machine Learning Research	AWS Machine Learning Blog	Google DeepMind Blog	Google Research Blog	KDnuggets	Machine Learning Mastery
MarkTechPost	MIT News – Artificial Intelligence	OpenAI Blog	The Berkeley Artificial Intelligence Research Blog	TopBots	Towards AI