Generative AI (GenAI) can offer many benefits for students, educators, and researchers alike. However, when deciding whether to use GenAI tools, consider whether their intended use is ethical and permitted.
To critically and carefully use GenAI,
As GenAI makes its way into every facet of our daily lives, you may find it tempting and convenient to incorporate it into your academic life. However, when considering GenAI to study or prepare for assessments, it is crucial to exercise caution, adhere to institutional and instructor guidelines, and consider whether the tool is helping you to learn or leading to learning loss.
While GenAI can be a powerful tool for you when seeking to understand abstract and complex concepts, learn to recognize its limitations and avoid becoming overly reliant on it. By striking a balance between traditional study methods and, if permitted, thoughtful integration of GenAI into study routines, you can make informed decisions that enrich your learning experience while maintaining academic integrity.
GuidelinesBe mindful of several factors before using GenAI as a tool to study or for assessment preparation:
|
 |
| Capabilities | Limitations |
| Personalized Learning: Adapts study materials and learning paths to individual needs and preferences, including as assistive technology. | Accuracy: AI-generated content may contain errors, leading to misunderstandings or misinformation. |
| Interactive Learning: Provides engaging, interactive experiences through quizzes, simulations, and multimedia. | Bias: Study materials may reflect biases present in the AI's training data, potentially perpetuating stereotypes. |
| Immediate Feedback: Offers instant feedback on exercises and assessments, helping to reinforce learning. | Dependence: Overreliance on AI tools may reduce the development of independent study skills and critical thinking. |
| 24/7 Availability: Accessible at any time, providing support outside of regular classroom hours. | Contextual Understanding: AI may struggle with nuanced or complex subjects that require deep contextual understanding. |
| Resource Variety: Generates diverse study resources, including summaries, flashcards, and practice questions. | Ethical Concerns: AI tools may not adhere to privacy and data protection guidelines, especially when handling personal information. |
| Language Support: Assists with language learning and translation, making studying more accessible for non-native speakers. | Customization Limitations: The use of AI may not align perfectly with specific curriculum standards or individual study needs. |
Generative AI (GenAI) can be used to automate numerous teaching tasks. For example, it can generate slide decks and lesson plans, build assessments, and provide personalized feedback to students. This automation can relieve educators from the burden of manually completing these tasks, allowing them to focus on more strategic aspects of teaching, such as facilitating discussions, designing engaging learning activities, and mentoring students. By reducing the cognitive load associated with routine tasks, GenAI can enable educators to allocate their time and energy more effectively.
GuidelinesEducators can leverage the advantages of GenAI and maintain a cautious approach to ensure a safe, transparent, and ethical learning environment. Before using GenAI, educators must ensure compliance with institutional policies, acknowledge the potential biases and inaccuracies in AI outputs, protect students' intellectual property rights, and exercise caution when using AI as a grading tool. Example, AI detectors, tools used to identify AI-generated content, might not be approved nor recommended for use as they can compromise student privacy, and they often produce false positives, which can result in wrongful accusations of misconduct. UNESCO AI competency framework for teachers "empowers them to use the technological tools in their teaching practices in a safe, effective and ethical manner". |
 |
| Capabilities | Limitations |
| Personalized Learning: Can tailor educational content to individual student needs and learning styles. | Accuracy: AI-generated content may contain factual inaccuracies that can misinform students. |
| Scalability: Provides educational support to a large number of students simultaneously, enhancing access to learning resources. | Bias: Teaching materials may reflect biases present in the training data, potentially perpetuating stereotypes. |
| Engagement: Creates interactive and engaging learning experiences through multimedia content such as animations, quizzes, and simulations. | Contextual Understanding: AI may lack the depth of contextual understanding needed for nuanced teaching and learning. |
| Assessment and Feedback: Can quickly assess student work and provide immediate feedback, promoting faster learning cycles. |
Ethical Concerns: Issues related to student data privacy and consent must be carefully managed. Do not enter student work into a third-party AI tool without proper consent. |
| Accessibility: Improves accessibility for students with disabilities through features like text-to-speech, language translation, and more. | Dependence: Overreliance on AI tools may reduce students' development of critical thinking and problem-solving skills in students. |
| Resource Creation: Assists educators in creating diverse teaching materials, lesson plans, and activities. | Customization Limitations: The use of AI may not align perfectly with specific curriculum standards or individual study needs. |
While generative AI (GenAI) can serve as a search tool and generate coherent responses to simple prompts, it presents significant challenges when it comes to accuracy, bias, ethics, data privacy, and source verification.
Traditional search engines offer direct links to sources, enabling users to verify information. In contrast, AI-generated responses often lack clear citations, making it difficult to trace the origin of information or confirm its accuracy. Overcoming these challenges is important in determining whether GenAI can and should be used as an effective and reliable search tool.
Be mindful that GenAI is not a suitable tool when searching for academic literature.
GuidelinesUnderstanding GenAI Outputs: GenAI tools are trained on vast datasets from both published and unpublished sources, using pattern recognition to generate the 'best output' in response to text-based prompts. Outputs can be in the form of text, images, audio, or video, and their quality depends on the user's prompt and the undisclosed data sources used for training. These factors contribute to significant limitations when using GenAI for information retrieval. Accuracy and Verification: GenAI may produce convincing but incorrect or misleading information, a phenomenon known as 'hallucination.' To ensure accuracy, it is essential to cross-reference AI-generated content with reliable sources and fact-check information to avoid misinformation. Ethical and Legal Considerations: The inclusion of copyrighted material in training data raises legal and ethical concerns. Users should be aware of potential copyright infringements and ensure that AI-generated content is used ethically and responsibly. |
 |
|
Prompt Creation: While GenAI itself can be used to generate information, it can also assist in creating effective search prompts. By using GenAI to refine and optimize search queries, users can improve the relevance and quality of the information retrieved from traditional search engines. Choosing the Right GenAI Tool: Some GenAI tools provide source citations and clear references for their outputs. Select tools, such as Perplexity, that offer transparency in their data sources, allowing users to verify information and ensure its reliability. Fact-Checking and Cross-Referencing: To overcome the limitations of GenAI, you must cross-reference AI-generated content with reliable sources. This practice helps to ensure accuracy; prevent plagiarism or copyright infringement; and avoid the spread of misinformation. |
|
| Capabilities | Limitations |
| Efficiency: Quickly retrieves large amounts of information, saving time and effort. | Accuracy: AI-generated information may be factually inaccurate or misleading. |
| Natural Language Processing: Understands and responds to queries in natural language, making it user-friendly. | Bias: Responses may reflect biases present in the training data, leading to skewed or partial information. |
| Comprehensive Search: Can pull information from a vast range of sources, providing broad overviews. | Transparency: Lack of clarity on data sources can make it difficult to verify the reliability of information. |
| Customizable Queries: Allows for tailored searches based on specific needs and contexts. | Ethical Concerns: Potential misuse of copyrighted material and ethical issues related to data privacy. |
| Multi-modal Retrieval: Capable of retrieving diverse types of content, including text, images, and videos. | Hallucinations: Outputs may contain convincing but false information, known as hallucinations, that require careful evaluation. |
| Accessibility: Enhances access to information for users with varying levels of expertise and from different fields. | Verification: Users must cross-reference AI-generated information with reliable sources to ensure accuracy. |
Generative AI (GenAI) can be used in research to enhance data analysis, streamline research processes, and facilitate discoveries. However, researchers must ensure that its use aligns with ethical guidelines, institutional policies, and research objectives and that research practices align with standards of research integrity and methodological soundness.
GuidelinesGenAI tools can support research activities when their specific applications and extent of use are appropriate, approved, and adhere to departmental or university policies. The use of AI in research is governed by the same policies and regulations as non-AI-assisted research. Use it responsibly by upholding principles of academic integrity and remaining mindful of ethical considerations. Depending on the stage of research, additional policies may become relevant to determine the acceptability and extent of GenAI use. For example, when disseminating research, check the policies of your target journal or publisher. If seeking funding, be aware of the funder’s policies regarding the use of GenAI tools for developing grant proposals. |
 |
In research, effectively employing Generative AI involves carefully assessing the tool's compatibility with objectives, methodology, data sources, and ethical standards while prioritizing transparency, validation, accuracy, and security.
Generative AI (GenAI) can be used for various writing tasks, such as organizing articles and reports, blogging, or brainstorming. While GenAI tools have versatility and potential to assist you in the writing process, it's important to ensure you have permission to use them.
GuidelinesBefore using any GenAI writing tools for coursework, verify whether their use is permitted by checking assignment instructions, noting any in-class guidance, and reviewing the course syllabus. If you are writing for publication, carefully read publication agreements and submission guidelines for references to AI. If none can be found, contact the publisher for more information. If, after reviewing instructions and confirming permission, you are unsure about acceptable uses of GenAI, ask your instructor or supervisor for clarification. Consider asking for guidance in writing (such as in an email) to ensure you are not missing any details. Note that what is allowed in one class, a section of a class, or even in an individual assessment in one class, may not be allowed in another class, section, or assessment. If you are using GenAI in a group project, check in with your group members before using GenAI. |
 |
| Capabilities | Limitations |
| Generating ideas / supporting brainstorming: GenAI can spark ideas or suggestions in response to aprompt, helping a user get "unstuck" on starting an assignment. It can also be prompted to argue or present alternative viewpoints and perspectives. For example, a student might input, 'Analyze the theme of identity in The Great Gatsby,' and the AI can generate ideas such as the contrast between Gatsby and Nick's identities, or the influence of social class on personal identity in the novel. | Handling Complexity: GenAI can generate errors, inconsistencies, and oversimplifications when used to assist with academic work. This can be seen in how it manages i) complex texts (e.g., nuanced texts, intricate reasoning); ii) subjects that span more than one discipline, or involve interconnected theories; iii) inputs requiring the AI to help with the synthesis of information or critical thinking; and iv) maintenance of context, particularly in longer interactions. |
| Aiding in paper structuring and drafting: GenAI can assist in organizing thoughts, creating outlines, and drafting with commonly-used disciplinary vocabulary and phrases (note: not appropriate in all disciplines or contexts). GenAI can be used to reverse outline work, too. | Summarizing and paraphrasing: GenAI does not always create a fully accurate summary or paraphrase, which can lead to the omission or misrepresentation of important details. Concerningly, GenAI also tends to patchwork plagiarize. |
| Demonstrating genre conventions: GenAI can be used to better understand the conventions and styles typical of various genres. For example, it can generate persuasive essays with strong arguments, narrative essays with personal stories, or descriptive essays with vivid details, helping students recognize and analyze each type. | Generating biased and inaccurate information: GenAI tools do not reliably generate unbiased, balanced, and accurate information, and their outputs require human verification. The user must have critical thinking skills and the ability to fact-check and verify outputs. |
| Answering grammar questions: Many use GenAI to revise, edit, and proofread their writing (examples). This can lead to the ability to identify redundancies; idiomatic expressions (e.g., “Can you list some idioms that express surprise or astonishment?”) or synonyms (e.g., "what is another word for 'happy'?"). GenAI can also help writers with punctuation and sentence structure, leading to more polished work. Be sure to see the "Limitations" column for important issues around editing and proofreading. | Inconsistently indentifying grammatical errors: GenAI tools do not consistently identify grammatical errors, especially in complex texts. Better outputs require better inputs (prompts), meaning that you require some grammar knowledge to be specific in your prompt (e.g., "Check for unclear pronoun references"). As AI tools become more integrated into educational platforms and software (e.g., GrammarlyGo in Grammarly), you might unknowingly use AI-assisted features, even when their use is unauthorized. |
| Translating text and recommending vocabulary: Writers can use GenAI to translate text into different languages and choose vocabulary for their writing. Make sure that you verify and learn from outputs rather than having the AI do the work for you. | Translating accurately: These tools do not always produce fully accurate translations, particularly for your intended meaning or context, potentially resulting in misunderstandings or miscommunication. They can curtail learning if used without verification of outputs. Use translation functions only if permitted. |
| Improving accessibility: These tools can provide real-time, inexpensive or free writing assistance for writers with dyslexia, Autism or ADHD. For example, the prompt "I have a 20-page research paper due in six weeks. Can you help me create a plan to complete it on time?" may help a student with ADHD (or any student, for that matter) begin and manage the writing process. | Requiring User Expertise: Effective use of GenAI requires that you have the basic subject expertise, critical thinking, creative thinking, and revision skills to manage and correct limitations. |
| Levelling the playing field for scholars: GenAI can improve the chances for undergraduate or graduate students who communicate in a dialect of English, or whose first language is not English, to publish their knowledge (Tai, et al., 2023, p. T13). Care must be taken to check outputs for accuracy, review publication agreements and assess for plagiarism (GenAI can patchwork plagiarize, for example). | Generating a uniform style, linguistic uniformity: Cultivate and maintain your own writing style, maintaining your authorial voice. GenAI tools tend to default to Standard American English (SAE) and a uniform, generic style. The MLA-CCCC Joint Task Force on Writing and AI warns, “[s]tudents may face increased linguistic injustice because LLMs promote an uncritical normative reproduction of standardized English...that aligns with dominant racial and economic power structures” (2023, p. 7). For example, by correcting someone’s written dialect of English to SAE, AI tools can perpetuate the loss of cultural diversity and identity. However, some emerging AI tools (such as Latimer) have been trained on culturally and linguistically diverse data. |
Most reputable academic presses and publishers have clear guidelines in place that communicate best practices and appropriate use of generative AI (GenAI). As such, authors must review guidance in submission guidelines and publication agreements before using this technology in research or when preparing manuscripts or articles. Such direction is often featured on publishers' websites under “Author Guidelines.”
GuidelinesWhile each publication may have slightly different requirements, they typically emphasize transparency regarding the use of artificial intelligence. Conventional guidelines require that authors
In academic and professional writing, authors are responsible for ensuring the accuracy, integrity, and originality of their work. This responsibility extends to fact-checking, proper citation of sources, ethical handling of data, and taking ownership of the ideas presented. Although GenAI can generate text, it cannot guarantee accuracy, cite sources, maintain ethical standards, or claim ideas as its own. Therefore, it cannot be credited with authorship. |
|
Authorship ConsiderationsIf co-authoring a paper, be sure to discuss with collaborators in advance whether GenAI will be used as part of the research or writing process, and in what specific ways. Given that authorship requires the writer to take full responsibility for the accuracy, integrity, and originality of the work, it is not appropriate to identify GenAI as an author or co-author of a manuscript or paper. As the COPE Committee on Publication Ethics notes, "AI tools cannot meet the requirements for authorship as they cannot take responsibility for the submitted work. As non-legal entities, they cannot assert the presence or absence of conflicts of interest nor manage copyright and license agreements" (2023). Image ConsiderationsSome publishers currently do not permit the use of GenAI images due to copyright and research integrity concerns. Before submitting their work, authors are responsible for verifying with a publisher whether they can use GenAI images in their manuscript or article. Review ProcessUnless given explicit permission to do so by academic publishers, conference organizers, or funding bodies, AI should not be used during the peer review process to critique the work of other scholars. Work-in-review is not considered public or fair use and should not be uploaded to AI platforms, as doing so can constitute a violation of intellectual property. Within academic publishing, there are concerns that GenAI is contributing to ‘junk science’ and accelerating submissions of fraudulent research produced by paper mills. For context, in May 2024, Wiley shut down nearly twenty journals that the company purchased in 2021 from Hindawi, partly in response to high retraction rates (more than 11,300 articles) associated with work produced by paper mills. Wiley also announced its adoption of a new detection software as a means to combat paper mills and detect articles produced by GenAI. |
|
Animations, images, maps, music, presentation slides, voiceovers, and, increasingly, videos, among other multimedia products, can be created with GenAI. However, it is imperative to verify that one has permission to use these technologies, whether from an instructor, research supervisor, collaborator, or publisher. Be aware that due to pending court cases, there are risks associated with using AI-generated multimedia, particularly visuals.
In audio and visual content creation, GenAI tools offer diverse capabilities. GenAI tools can generate audio and images from text prompts, providing a wide range of realistic and stylized outputs. Users can also edit existing images, adjusting colours, features, and elements to enhance quality or creativity. GenAI tools can create custom graphics and animations for videos or presentations, potentially improving visual appeal and engagement.
However, users must be aware of the technology's limitations, such as potential inaccuracies in content, copyright infringement risks, biases in training data affecting outputs, deepfake threats, and violation of publication policies.
GuidelinesFor students: Before using GenAI for multimedia purposes, verify whether permitted by checking the assessment instructions or course syllabus. If the rules regarding GenAI use are unclear or ambiguous, ask specific questions about its use and clarify when and where it is allowed, particularly for assessments. For Instructors: To help students understand the appropriate use of GenAI, establish and communicate transparent policies regarding AI use in the assessment instructions and course syllabus. Clearly state the potential benefits and limitations of using GenAI as a tool to create multimedia content, emphasizing the importance of maintaining academic integrity, originality, and artistic expression. |
 |
| Capabilities | Limitations |
| Efficiency: Quick generation of high-quality content, saving time and effort compared to traditional methods. | Inaccuracies: Audiovisual outputs produced by GenAI tools may depict incorrect information. |
| Automation: Access to innovative and creative outputs that might be difficult to achieve manually. | Copyright Issues: GenAI-generated content may contain& copyrighted information, raising legal concerns. |
| Customization: Ability to tailor content to specific needs and preferences, ensuring alignment with the user's vision. | Bias: GenAI tools may generate biased or discriminatory images, reflecting the biases present in the training data. |
| Support: Assistance throughout various stages of the creative process, from brainstorming to final editing. | Publication Policies: Generated content may violate publication policies. Publishing AI-generated visuals is currently not recommended due to pending copyright cases. |
| Enhancement: Offers design suggestions and improvements, potentially optimizing the quality and appeal of the content. | Deepfakes: GenAI can create convincing deepfakes, leading to potential fraud, privacy, and exploitation concerns. |
If choosing to use GenAI tools in your creative process, keep in mind these important considerations:
Note: View GenAI as a supplement to rather than a replacement for your unique ideas and talents. Above all, focus on cultivating your originality, innovation, and self-expression, as these aspects of your work are invaluable and cannot be replicated by any AI tool.
