Student Perceptions of AI: Recommendations for Schools and Teachers
Researchers at foundry10 offer recommendations for schools and teachers regarding AI use in the classroom.
This is Part Three of our three-part series on high school students’ perceptions of AI. See Part One and Part Two for more information.
To better understand student use and perceptions of AI, foundry10’s Digital Technologies and Education Lab conducted focus groups with 33 high school students from across the U.S. Based on the insights and concerns shared by students involved in the study, here are our top recommendations for educators.
Recommendations for Educators and School Districts
Teach students how to use AI responsibly, including the limitations of large language models.
Students in our focus groups showed some AI literacy, but most of what they learned came from trial and error, social media, or their peers. Most students formed the basic impression that tools like ChatGPT compile information from the internet and respond to queries, but they had limited awareness of its functionality and limitations.
We had a thing in the beginning of the year where our English teacher went over acceptable ways to use AI that can help you without taking away from you thinking for yourself in unacceptable ways. It’s not like it’s encouraged to cheat in our school, but it’s like to use it to your advantage in a smart way. — Savannah (age 16, grade 11, Focus Group 1)
Create opportunities for students to learn how to prompt engineering skills, identify misinformation, and avoid common pitfalls.
These skills will help them use AI responsibly and better prepare them for a world in which generative AI is rapidly changing how we work and learn. Several valuable resources exist to help teachers deliver quality AI literacy education: see MIT Media Lab’s Ethics of AI Curriculum, Stanford’s AI Resources dashboard, and the International Society for Technology Education (ISTE)’s AI resources and classroom guide on ethics in AI.
Focus on incorporating AI as a learning tool and steer clear of outright bans.
Many students reported that their schools or districts had banned AI entirely. These total bans came with increased surveillance and restrictions on other forms of technology (e.g., no computers in class).
I do think they’re going to have to incorporate it because it is getting so big, that if they try to just ban it completely, kids are still going to find a way to use it. So I think if they find a good way to incorporate it into classes and make it a learning tool, it would be good. — Jessica (age 16, grade 11, Focus Group 4)
These measures may have adverse impacts on students and seemed to have limited success. Several students in our study who came from total-ban schools still reported regularly using AI for their assignments.
We recommend moving away from total bans and working instead to integrate AI, when appropriate, as a valuable learning tool. Teachers can use resources like this toolkit and those linked above to make AI a helper rather than an adversary in the classroom.
Approach AI detection tools with caution.
While these technologies may improve, AI detection tools are error-prone and unreliable. These errors have led to lawsuits over false allegations of plagiarism. Teachers who use AI-detection tools should be careful not to interpret their results as conclusive evidence and should develop more holistic strategies for encouraging students to avoid plagiarism.
Help students build the skills to make ethical AI choices.
We recommend approaching student AI use with compassion and curiosity, seeking to understand when and why students use AI tools even when they are not allowed.
I heard about it, a friend of mine in a class, we had an optional extra credit essay. She used ChatGPT when it was first introduced, turned it in, got 100%, got the extra credit and I was like, wow, I should have done that. — Rachel (age 16, grade 11, Focus Group 2)
Our focus group findings suggest several strategies teachers and administrators can employ to set students up for success when it comes to ethical decision-making, including:
- Modeling specific examples of ethical use.
- Engaging students in the why behind assignments, clearly linking each task to learning goals and future success.
- Communicating the value of independent work and helping students identify areas where AI overreliance could harm their short- and long-term outcomes.
- Giving students clear next steps when tempted to use ChatGPT for cheating.
- Ensuring that the holistic expectations placed on high schoolers’ time and attention are within reasonable limits, reducing unnecessary time burdens where possible.
Provide clarity around policies and expectations for student AI use.
Students called for clarity on acceptable versus unacceptable AI use so they could explore ChatGPT within reasonable limits without worrying about misstepping. Defining and helping students uphold guidelines for responsible use should be a top priority, and expectations should be clarified (e.g., developing written guidelines, modeling specific examples of appropriate AI use, teaching students how to cite LLMs.)
Students also reflected on high variability in how their different teachers viewed AI. While class-by-class policies may be necessary because subject-appropriate use varies, school-wide expectation consistency may help prevent confusion.
Prioritize student voice in the development of policies and guidelines.
Students in our focus groups showed a remarkable capacity to engage with nuance around ethical and practical AI dilemmas, collaboratively design guidelines, and differentiate between more and less acceptable use of AI in school.
It would be good to see generative AI used as a society so that no one’s falling behind or no one’s advancing too much and we’re all just using it together… Which is also why I think that guidance on how to use ChatGPT or generative AI in general in education would be super helpful for students. — Aubrey (age 17, grade 12, Focus Group 3)
Involve students in co-creating guidelines for AI use.
Integrating student voice helps increase student buy-in and could provide unique perspectives. Schools and districts should solicit students’ opinions in facilitated discussions and integrate their recommendations into policy decisions. Teachers should collaboratively create expectations around AI use with their classes at the beginning of the semester or year and revisit these expectations as new assignments emerge to engage students in conversations about what responsible use looks like in different contexts.
I feel like in order to be a responsible user, use it to not blatantly cheat and use it for every single assignment, but use it here and there for ideas or inspiration about certain things and not just abuse it, I guess. — Joseph (age 17, grade 12, Focus Group 2)
Support teachers in learning and teaching AI.
I don’t know any teachers [who use ChatGPT]. Most of mine, they don’t even know how to run student view, so they don’t really know what they’re doing with it. — Rachel (age 16, grade 11, Focus Group 2)
Teachers need targeted professional development and district-level support to leverage the benefits of AI and tools like ChatGPT. District administrators can support teachers in three primary ways:
- Providing consistent and clear guidelines for balancing ChatGPT use with academic integrity concerns
- Providing guidelines and standards of use for teachers that align with district student data privacy policies
- Investing in AI training programs that will help prepare teachers to maximize the educational uses of AI.
Many of these recommendations parallel those that came from the findings of our Teacher Perceptions of AI study, highlighting the shared needs of educators and students in a rapidly evolving space.
Looking for additional resources? Check out our evidence-based resources guide for educators. Read more about the Digital Technologies and Education Lab’s work in this area.
