����ֱ��

_{The image above was generated on Oct. 12, 2023, with Canva’s Magic Design tool using the prompt “abstract image of a graph with scattered data points converging into a line.}

Generative AI works by training a statistical model to identify and "learn" patterns within a large dataset. When prompted, it generates new outputs by predicting what is most probable based on those detected patterns.

For instance, imagine a generative AI model trained on all the text messages you’ve ever sent. If the model finds that the word “thank” is most often followed by “you” (as in “thank you”) rather than “me” (as in “thank me”), it will prioritize that pattern. So, if you prompt the AI to draft a thank-you message after meeting a friend for coffee, it will likely include the phrase “thank you” rather than “thank me.”

Want to dive deeper into how generative AI works? Check out this for more insights!

_{The image above was generated on Oct. 12, 2023, with Canva’s Magic Design tool using the prompt “colorful abstract image of computer with scattered data, tools, clouds.}

The following is a list of generative AI tools with various functions for teaching, learning, and research. When exploring these tools, we encourage you to consider how generative AI may harvest and share data, copyright issues, and accessibility considerations and to note that some platforms could transition into subscription-based or pay-to-use services. This list is not exhaustive, as this new technology is part of the ever-changing digital landscape. If you have an AI tool that you have been using for teaching and learning, send us an email and let us know!

_{The image above was generated on Oct. 12, 2023, with Canva’s Magic Design tool using the prompt “abstract image of college students using a computer cooperating.}

The U.S. Department of Education has suggested that we think of AI as “more like an electric bike and less like robot vacuums” (2023). Drawing on this metaphor, we encourage you to think of AI as a technology that can be leveraged to support, rather than undermine, the learning process. As is the case with any technology, however, using generative AI effectively in the classroom requires forethought and intentionality. If you are interested in incorporating AI into your own courses, we recommend that you begin by reflecting on your course learning objectives. This will inform whether and how you incorporate AI into course assignments.

_{The image above was generated on Oct. 12, 2023, with Canva’s Magic Design tool using the prompt “abstract image of college students using a computer cooperating.}

Although generative AI is a powerful tool, it has several significant limitations. Namely, generative AI sometimes “hallucinates” inaccurate information, including attributing content to the wrong source or fabricated sources. It can also be used to create “deepfakes”–images, videos, or audio of a person doing or saying something that they did not actually do (e.g., see ). Furthermore, generative AI has been shown to reproduce non-inclusive language and biases, including gender, racial, anti-Muslim, disability-related, and other biases present in the data on which the statistical model was trained (e.g., Sun et al., 2023; Johnson et al., 2022; Abid et al., 2021; Whittaker et al., 2019). Lastly, generative AI raises a host of concerns around data privacy and copyrights. 

As educators, we can provide students with guidance on how to use AI responsibly and ethically, as well as opportunities to practice essential AI literacy skills. This includes skills like fact-checking, verifying AI-generated outputs, and writing effective prompts.   

_{The image above was generated on Oct. 12, 2023, with Canva’s Magic Design tool using the prompt “abstract image college student cheating on exam.}

The rapid gain in popularity of generative AI tools has raised concerns among educators that some students will misuse these tools in educational settings (e.g., use a chatbot to compose an essay or get the answers to a closed-book exam). As part of equity-minded assessment, we encourage educators to be explicit and transparent regarding which uses of AI, if any, are acceptable, as well as how students should cite AI-generated content. You can include this information in the instructions for individual assignments or a more general statement in your syllabus (e.g., Guidance on crafting an AI syllabus statement, ). Furthermore, for which it is not permitted. For students, we have also developed a list of .

Lastly, although we empathize with concerns around AI and academic dishonesty, it is important to recognize that and prone to error and biases. For example, emerging evidence suggests that GPT detectors are biased toward misclassifying non-native English writing as AI-generated (). Learn more about !