The purpose of this guideline document is to facilitate the responsible use of Generative Artificial Intelligence (GenAI) tools within the Bilkent community. This document outlines the definition of GenAI, provides illustrative examples, discusses potential use cases and risks associated with its use, and offers recommendations for its appropriate applications.
This document has been created by reviewing the literature, conducting surveys with both students and faculty members, and organizing focus group discussions exclusively with the students. It will be updated regularly with the latest information and insights to match the rapidly changing technology.
Introduction:
Bilkent University adopts a progressive and anticipatory approach toward using GenAI tools as a beneficial and imaginative influence in the field of education, integrating these tools into creative teaching/learning methods and evaluation. While integrating novel technologies in education, Bilkent University maintains the core value that students must maintain academic honesty in all academic work. The utilization of GenAI in all academic work is subject to the regulations outlined in the academic integrity policy.
What is GenAI? What are GenAI tools?
Generative AI is a subfield of artificial intelligence that creates new data or content instead of just analyzing existing data. This data or content can vary across text, code, images, videos, music, or even a blend of these formats. GenAI operates through generative models, like Large Language Models (LLMs), which utilize extensive datasets to create responses based on given queries or prompts. Unlike traditional AI systems that rely on predefined rules, GenAI uses machine learning algorithms, particularly deep neural networks, to learn patterns from large datasets and produce new content that resembles the input data.
Some examples of GenAI tools are ChatGPT, Bing, Bard, Dall-E, Perplexity.ai, Jasper, Midjourney, Amper, Dadabots, CodeStarter, Codex, Podcast.ai, Stability.ai, Adobe Firefly, Runway ML, Canva, SlidesAI, and SlidesGP.
Use cases of GenAI in higher education
GenAI presents valuable opportunities to streamline time-consuming tasks, such as data processing and background information gathering. Recognizing the potential of this technology to augment efficiency is essential. As GenAI technologies rapidly integrate with the educational landscape, it is vital to recognize that their applicability might vary across different disciplines, courses, or projects.
Students are advised to check with specific course syllabi and instructors regarding the permissibility and appropriate methods for employing GenAI tools in specific courses / for specific assignments.
Below are some quotations from Bilkent Faculty members and students on how they currently use or are planning to use GenAI tools.
From faculty members
“The students enrolled in this course are required to use AI tools during the lab and when studying at home, specifically when solving self-study tasks, to address problems related to formulas and functions (e.g., using the wrong delimiter) that they have generated. Additionally, they are expected to utilize appropriate AI tools for correcting or updating erroneous code when implementing programming tasks related to VBA and Python.”
“I have shown students how to use GenAI for speaking purposes and make it their conversational partner.”
“Students have used GenAI to demonstrate its strengths and weaknesses in writing. GenAI has proved to be an effective tool for idea generation and inquiry. It is awful for analysis and finding reputable sources that exist.”
“I had students generate drafts of tasks by using a GenAI tool. This helped me understand what they do with it. It allowed me to discuss with them the ethical implications and the strengths/weaknesses of the tool.”
“I have demonstrated to students what GenAI can do in responding to my assignment prompts, where it falls short of the assignment expectations, and why. This fosters good discussions about how to make an argument in academic writing because GenAI currently is not good at making arguments and instead reports on general information. This helps students understand the difference between making and defending a stance and summarizing the general issue.”
“On Moodle forum of my course, I have asked the students to answer the sample questions that I was sharing with them by using LLM tools (e.g., ChatGPT) and discuss the answers that the tool was generating and evaluate them.”
“I use GenAI to create context and exercises to practice some grammar topics, vocabulary, reading, and writing skills.”
“I have been using AI to generate discussion questions and design the end product in my lessons. For example, If I ask my students to design a poster on the impacts of climate change, I resort to AI to give me possible subheadings for the poster.”
“Students will be given several GenAI prompts to execute and summarize the responses. The prompts were to emphasize how and why GenAI should not be used in certain scenarios (completing course work vs. supporting learning) and also to emphasize the ways that GenAI can be used to support learning (suggesting learning strategies, resources, explaining concepts).”
“The students will be asked to use GenAI to generate essay responses to a prompt and have students analyze them for content and ideas (an area where students particularly struggle with).”
“I am planning to ask my students to use proper prompts (help them to learn about prompting) at GenAI to get specific cases for the “Instructional Design” course. They may also prepare lesson plans by using GenAI and check its properties according to certain criteria (again improve prompting skills).”
From students
“I like to give a relevant code to ChatGPT and ask it to explain the code step by step better to understand the process of executions in a certain program. Which has proven a valuable resource.”
“I don’t see myself using GenAI for learning, but I would use it for communication (e.g., generating an e-mail template) or other tasks that I would otherwise waste a lot of time on due to anxiety.”
“Summarizing topics for me, creating study plans, asking it to explain the parts I don’t understand.”
“To better understand concepts as ChatGPT is a learned tutor who is available for you 24/7”.
“For example, I sometimes used it to find other ways to solve problems or get a brief summary of topics.”
“I usually give it a piece of text and make it come up with comprehensive questions. After that, I want a brief summary of the text from GenAI. I read and came up with answers. I prompt my answers to GenAI and want it to correct them. After all this, I read the text I supplied, checked all the answers, and came up with additional questions. Then iterate once again. It’s a good learning partner.”
“I will create personalized syllabi, get help with self-organization and take notes more effectively.”
Risks associated with using GenAI tools
While GenAI tools may offer potential benefits, they also come with inherent risks. It is essential to consider the following points seriously:
● Hallucinations: GenAI may involve integrating loosely relevant facts and data to create factually incorrect statements or conclusions, as well as the generation of entirely nonexistent personalities, institutions, phenomena, and data.
● Security flaws: GenAI might cause problems with security issues, which involve disclosing information concerning individuals, institutions, processes, or corporations.
● Privacy issues: GenAI may intensify the complexity of privacy issues, which involve the unauthorized disclosure of personal, private, or corporate information, including intellectual property, to third parties without proper consent.
● Offensive content: GenAI may involve creating and distributing text, images, sounds, or videos that can be politically, socially, sexually, religiously, or culturally offensive without obtaining consent from relevant parties.
● Fake citations and references: GenAI may generate fictitious sources of data or facts, leading to inaccurate or misleading information.
● Inaccuracy: GenAI may contribute to inaccuracy, which is creating and disseminating content or information that is inaccurate or outdated, thereby contributing to misinformation.
● The risk of plagiarism: GenAI may cause the risk of plagiarism, which arises from the absence of proper citations and a comprehensive understanding of the extent of possible replication from existing resources.
● Bias: GenAI may have bias, which occurs when content is generated based on predominantly existing opinions, data, or views while ignoring other critical, accurate, or less frequently discussed perspectives.
● Relying solely on AI tools as a source of information may result in missing out on alternative, more accurate, and valuable ideas, data, and resources, destructing the critical thinking process.
● Instructors being unclear about what is acceptable and what is not: Plagiarism issues may be generated due to poorly defined rules and regulations by instructors specific to each course resulting in students’ lack of clear understanding of such policies.
● Faculty members having varied understandings and practices of AI use: Letting students get help from AI at different levels, i.e., one faculty totally letting students use AI tools to write a portion of their assignment, but another faculty banning such tools in multi-section courses offered by several instructors.
● Not being able to do anything without it: Loss of self-confidence by students, breaking up the learning-by-doing process.
Responsible use of GenAI tools and recommendations
Faculty members are recommended:
To include a statement in their course syllabi indicating whether they will
- allow students to use GenAI technologies in the course, and if so, to provide details on which usages are allowed and which are not,
- incorporate GenAI technologies into the course or not, and if so, to provide details on how they plan to do so.
Students are recommended:
To carefully review syllabi of all their courses about use of GenAI: Students must never use GenAI if it is not explicitly allowed in the syllabi by the professor.
To consult with their professors (thesis advisor and/or course instructor) about using GenAI: If a student is in doubt about use of GenAI in a specific course or other academic work, they must communicate with the corresponding faculty member. Their expertise and guidance will help students navigate through the ethical considerations and make informed decisions regarding integrating GenAI into their academic work. If they are still unsure about whether to use GenAI or not in a particular situation after reviewing the syllabus and consulting with their professor, they should not use to be on the safe side.
To cite other peoples’ ideas and acknowledge the use of AI tools & be transparent about the use of AI tools: If students use content generated by AI tools in their work, they must provide appropriate attribution to the AI system and any human contributors. They should not submit assignments which include AI-generated content without giving proper attribution to avoid academic misconduct. They should learn how to cite the content generated by AI tools correctly.
To keep records of how and why AI tools were used: Maintaining comprehensive records of the usage methods and rationales behind GenAI tools will contribute to transparency and accountability in students’ academic life.
To consult additional sources to verify factual accuracy: When using AI-generated information, it is crucial to critically assess its accuracy and credibility.
To approach GenAI tools critically: Students should regard GenAI tools as supportive aids to enhance their learning while still upholding their creative and critical thinking skills.
To avoid sharing personal, confidential, or sensitive information: Students must exercise caution while providing AI tools with personal data, ensuring appropriate consent whenever necessary.
To understand the biases and limitations of GenAI: Before integrating GenAI into assignments, it is important to thoroughly comprehend its biases and limitations.
To keep up to date about using GenAI: It is crucial to stay current with the advancements in AI ethics, and responsible AI use to make informed decisions about AI integration in academic work. Participate in discussions and workshops on AI ethics organized by the university for a better understanding of responsible AI use in higher education.