AI and Academic Integrity
Much has been written about the possibility of students using AI tools like ChatGPT to “cheat” on assignments. The first step in helping students make informed choices about how and when they use AI is to have an open and honest dialogue to help them understand when using AI is OK and when it is not.
Talking with students about AI
Students know that ChatGPT other generative AI tools exist, and they may have tried them out. Like many, they are curious about what these new tools can and can’t do. They may be wondering how these new tools might make their lives easier.
The first step is to acknowledge students’ curiosity by having conversations with them. Talk to students about AI tools, and whether or not it is OK to use those tools (for schoolwork or otherwise).
Most online AI tools (including Chat GPT) are not designed for children, and should never be used by children under the age of 13.
While students under 18 and above 13 may create accounts for personal use with parental consent, teachers should NOT encourage students (no matter their age) to create accounts to use ChatGPT or other online AI tools unless the tool has been reviewed by the school division. You can talk with your school’s IFT if you have found a tool that you are interested in using with students.
Use of AI detection
Since the release of ChatGPT, there have been numerous tools released that claim to be able to detect if text was generated by AI. Turnitin.com, a writing feedback and plagiarism detection tool, has recently released their own AI detection tools to complement their existing suite of writing and plagiarism detection tools for teachers and students.
LCPS provides access to Turnitin.com to all high school teachers and students, and AI detection is included in our district subscription.
While AI detection may be a useful tool in a teacher’s toolbox, it should be one part of a larger strategy to help students understand when it is OK to use AI and when it is not.
These AI detection tools are new, and do not accurately flag AI writing 100% of the time. There is also evidence to suggest that AI writing detectors, including Turnitin.com, may flag writing of non-native speakers at a higher rate than the writing of native English speakers.
Strategies for designing assignments
If you are designing assignments or assessments that include a writing component, consider how you can best support students to ensure that they are able to demonstrate mastery of concepts. If you expect that students might use AI to support their learning, how might you guide them to ensure that they are using AI in appropriate ways? Here are some strategies to consider:
- Consider building your assignment so students are supported through the stages of the writing process. Turning in outlines, drafts, and revisions along the way may help to ensure that students are producing original work.
- Use Turnitin to review drafts (not just finished products) and share the review results with your students so they can see when their writing might have similarity to something drafted by AI. There is evidence to suggest that AI detection is not completely accurate, and can even be biased against certain students. So, AI detection should never be used as the final determination of whether or not a student used generative AI.
- Consider giving students options for how they demonstrate what they know. Instead of creating a written essay, perhaps students could create a video, infographic, or other original work. Students will appreciate having choice and will be excited to show what they know in a way that taps into their strengths and interests.
How does AI fit into existing policies on grading and academic integrity?
The LCPS Assessment and Grading Policy (Policy 5030) describes students’ responsibilities with regard to academic integrity and honesty in assigned work. If a student chooses to use AI to generate original work, and the use of AI was not approved by the teacher, a student could be in violation of this policy. Schools will address those violations in a manner consistent with other types of academic integrity violations.