Guidance for the Use of Generative AI

ChatGPT is a “chatbot” developed by a private company called OpenAI. Users can enter question prompts and within seconds ChatGPT will produce text-based responses in the form of poems, essays, articles, letters and more. It can also create structured responses like tables, bulleted lists and quizzes. ChatGPT can provide translation and copy language style and structure. It can also be used to develop and debug programming code. New and expanded uses continue to be developed and launched. A similar tool called DALL-E uses AI to create art pieces, and other AI tools have been created or are in rapid development to do even more – music, animation, multimedia video, powerpoints – and the list goes on.

Yes, students have already been exploring and using it to support completion of their coursework. Some courses explicitly encourage the use of ChatGPT for assignments. However, one concern is that students may be using ChatGPT to draft responses to homework responses without learning the material, and this presents the challenges and opportunities for reflection on teaching and learning at this time.

CEILS Education Research Talk with Jess Gregg. Demo portion starts just after the 5 minute mark.

Revolutionizing Education with ChatGPT: How AI is Transforming the Way We Learn – Learning and Technology with Frank

Dig Deeper (article): How ChatGPT Works

Discuss the potential: Many of our students will go on to become leaders at organizations that utilize and/or develop new AI technologies. How will these tools support advancements in your field (medicine, science, art, music, humanities, health, and more)?

Prepare students for the future when they will work and interact with AI: This technology is likely to develop and become embedded in many parts of our lives. Preparing students to thoughtfully engage with it, co-create with it and be curious about and know how to interact with other technological developments as they occur.

Seize the opportunity to center the importance of critical thinking and digital literacy. Students will have the opportunity in the future to break the cycle of spreading disinformation, lack of journalistic integrity in news, and elevating accurate and factual research and scholarship. Emphasize the importance of digital literacy, research, and writing skills with students; connect students to library resources for research and writing. As educators, we have an obligation to help guide our students through many types of literacy, including digital media and AI literacy. UCLA’s WI+RE has created the Understanding Misinformation: A Lesson Plan Toolkit, that is geared towards educators who want to prepare students to learn about misinformation. We can get students to vet information like experts.

Lean in to talking to your students about how learning happens: Learning happens when actively engaging with the course material, through conversations and dialogue leading to deepening conceptual understanding.

Discuss Academic Integrity with Your Students:

One of the main concerns instructors have expressed is how to uphold academic integrity and prevent the misuse of tools like ChatGPT (intentional or not). These concerns include:

• Plagiarism (copying and pasting the response that the tool provides; running material through multiple AI generators to avoid detection)
• Lack of proper citation of sources
• Inaccurate, misleading, biased, false, or limited information in responses to question prompts. While some AI detectors have been recently developed, it is unclear how effective they will be long-term and early reports indicate that individuals can easily avoid detection through simple modifications to produced text.

The UCLA Student Conduct Code states, “Unless otherwise specified by the faculty member, all submissions, whether in draft or final form, to meet course requirements (including a paper, project, exam, computer program, oral presentation, or other work) must either be the Student’s own work, or must clearly acknowledge the source.” Unless an instructor indicates otherwise, the use of ChatGPT or other AI tools for course assignments is akin to receiving assistance from another person and raises the same concern that work is not the student’s own. Please communicate this to your students, and consider incorporating this language into your syllabus.

Teaching Assistants will be seeking guidance on how to discuss ChatGPT with students and what to do if they suspect submitted work may be AI generated. In addition to talking with your students, make space for conversations with your TAs and other instructional team members to explore this topic and co-construct guidelines.

The ultimate decision and responsibility for how to teach about AI and the establishment of or revision of course policies related to its use lies with the instructor.

View this collaborative Google Doc: AI Examples and Resources to see examples (including Syllabus Language and Policy Language) and resources curated by UCLA’s teaching and learning community.

Discuss the ethical issues and limitations  of AI

Facilitate discussions with your students on the impacts of spreading disinformation or biased information, lack of regulation of companies that develop these technologies, and other dangers. While students will likely still continue to use ChatGPT and other tools like it, it is crucial that our community has this shared understanding of both dangers and opportunities.

These technologies are not infallible and their accuracy is subject to a variety of factors, some listed below:

• Prone to filling in replies with incorrect data if there is not enough information available on a subject.
• Lack the ability to understand the context of a particular situation, which can result in inaccurate outputs.
• Large, uncurated datasets scraped from the internet are full of biased data that then informs the models.
• Data is collected from the past, it tends to have a regressive bias that fails to reflect the progress of social movements.

Our nation has yet to catch up to the regulation needed to prevent the potential for tremendous harm when false or biased information is taken as fact. Our community must continue to explore the value and innovation that can come from AI while simultaneously contributing to the dialog about these potential harms.

Share current examples of scholarly discussion on this topic.

As our own UCLA experts have shared with our instructional community during the recent UCLA Virtual Town Hall: What is ChatGPT and How Does it Relate to UCLA’s Academic Mission, there are concerns about the ethics and practices around tools like ChatGPT. Algorithms can and do replicate and produce biased, racist, sexist, etc. outputs, along with incorrect and/or misleading information.

Additional Examples:

• Educause – Special Report on Artificial Intelligence
• Noam Chomsky: The False Implications of ChatGPT
• Schools Must Embrace the Looming Disruption of ChatGPT

On Requiring the Use of ChatGPT: Creating an account to use ChatGPT requires sharing of personal information. Depending on context, the use of ChatGPT may also mean sharing student intellectual property or student education records with ChatGPT under their terms and conditions of use. Individual students may have legitimate concerns and therefore may be unwilling to create an account. Discuss these concerns and consider alternatives.

• If you will be requiring use of ChatGPT, consider making this explicit in the syllabus (for a related example of considerations related to privacy, see Privacy Tips for Your Syllabus).

Protecting student privacy as required by FERPA: Academic records, such as examinations and course assignments, are considered a student record and protected by FERPA. For example, ChatGPT should not be used to draft initial feedback on a student’s submitted essay that included their identifying information. Asking ChatGPT to respond to question prompts would not be a FERPA violation, as no student information is provided to ChatGPT.

ChatGPT is currently in the process of review through UCLA’s Third Party Risk Management to understand where there will be gaps in accessibility and security.

As with any emerging technology, ChatGPT may not always be accessible by individuals with disabilities. Open a conversation with the Center for Accessible Education (CAE) for ideas on exploring accessible alternatives. As always, we encourage faculty to use this sample syllabus language to direct a student toward CAE to discuss their options for accommodations and support. Requests for support should be directed to caeintake@saonet.ucla.edu or the student’s listed Disability Specialist on their accommodation letter.

As the technology evolves, there may be a cost to using it, so continuing to revisit your learning goals and activities with respect to access is a critical equity issue.