Guidance onCourse Statements and Policies

Incorporate clear and transparent statements about the role and implications of AI in your course(s). In drafting such statements, consider adopting compassionate (rather than punitive) language, helping students understand why and how AI tools will be incorporated into the context of your course. The syllabus and dedicated pages of your Canvas site are great places to share AI-related course policies.

Guidance onAuthentic Assignment and Assessment Design

Many of the assignment design strategies recommended for deterring dishonest use of AI are consistent with good assignment design principles in general. Students are less likely to turn to all forms of academic dishonesty when assignments are authentic (applied to real-life circumstances), transparent (clearly and openly explained), and scaffolded (broken down into smaller, well-supported steps).

It's important to be clear with students as to the purpose of an assignment. It may be wise to consider the act of writing an essay as a way for students to strengthen their understanding of a topic or improve skill in presentation. In that way, the use of authentic assessments may entail a variety of products beyond traditional essays.

Use of a variety of media, in both assignment prompts and student demonstration of learning is consistent with Universal Design for Learning principles while also deterring the use of AI.

Finally, assignments that actually incorporate AI (asking students, for example, to analyze an AI-generated essay) also build information literacy, including student ability to detect AI.

Guidance onSpaces for Inquiry and Exploration

Create opportunities for students to reflect on, discuss, and questionthe possibilities and limitations of AI tools in their disciplines, the workplace, the broader sociocultural context, and/or everyday events. Such spaces contribute not only to AI literacy but encourage critical analysis of AI tools.

Additionally, you can ask your students to submit a transcript of their interaction with many GenAI tools (particularly ChatGPT and Co-Pilot) as a means of showing how they interacted with the tool. This work can be cited in MLA and APA styles, among others.

Guidance onUse of AI Detection Tools

There are multiple tools that screen text for whether it has been created by generative AI tools. This includes TurnItIn’s AI report, which as of Fall 2024 is part of SMCCD’s license with TurnItIn. These tools are significantly more unreliable than plagiarism detection tools, both in terms of false positives and false negatives, and probably will remain so. Based on this unreliability, any use of automated AI screening tools should be used with extreme caution or not at all.

Automated screening tools could be helpful as a starting point for discussing a submission with a student, including checking for student comprehension of the written content.

When making a determination that a student has submitted AI generated content as their own work in violation of the course and/or assignment requirements, it is recommended that the faculty does not rely on an automated report as their only evidence. In general, it is recommended to place greater emphasis on assignment design rather than detection methods.

As an alternative to an AI detection tool, the course syllabus could include the following statement:

"Be aware that coursework will be subject to plagiarism and AI generated content screening software, such as TurnItIn and ZeroGPT. If I believe that your submission has been generated by an AI tool without attribution, I might request you to meet with me in order to demonstrate comprehension of the material in order to earn credit on the assignment. Other possibilities include having to redo the assignment or not earning any credit on the assignment."