The research is in: Usage of AI chatbots and tools in universities and high schools is booming. Students and learners are jumping to these tools because they make their lives easier, and teachers are doing it too.

Despite making work easier for everyone, in education AI usage poses one massive problem, a problem that has been identified in numerous pieces of research. It makes students engage their critical thinking less because they can simply have the machine answer for them. This translates to poorer all-around education for everyone and is a major threat to how education works in the modern world.

What would have been a week’s worth of essay writing and research now becomes a day’s for the unscrupulous student who just wants to pass. In the past, we wrote about how to honestly use ChatGPT to help you in your research in a way that saves you a lot of time, but this will only help students who want to play it honest.

The AI in education problem

For everyone else, teachers and educators have been wracking their brains, thinking about how to create assignments that are AI-proof.

Perhaps they could use different wording for answers, or ask students to use sources from before a time when ChatGPT collects information from the internet, such as pre-2021.

Should you ban the use of these websites across school WiFi networks?

Should you use software to see if the text is AI-generated or human-made?

The short answer is that none of these are foolproof methods and that there is simply nothing you can do to stop students and learners from turning to AI completely.

Instead, as argued by two lecturers at the University of Sydney, Australia, teachers need to stop wanting students to go away from AI and start teaching and guiding them to use AI properly and honestly.

How to make AI-proof assignments? Motivate students to be more honest

According to Danny Liu and Adam Bridgeman of the University of Sydney, one of the best ways to get your students to use AI honestly is to motivate them to do so.

“Students are more likely to undertake assessment with integrity if they feel they have real choice about topic and mode, and seeing how the assessment meaningfully connects with their life and career,” they explain.

If they feel they are “being supported to build confidence and skills gradually” and feel “connected to teachers and peers and that they matter.”

Indeed, turning to AI to cheat is a product of stress. If students feel they are stressed and need to hand in the assignment, they will be more likely to use ChatGPT dishonestly.

“Assessments with clear instructions and criteria, have meaningful and appropriate challenge, that provide sufficient time for completion, and which help students develop confidence in their abilities (e.g. through structured drafts and feedback) will lead to more positive academic integrity outcomes,” the pair write.

Aside from psychological changes, assignments themselves need to be redesigned in this modern world to mitigate the risk of generative AI, but also influence students to use AI honestly in a way they can learn from.

The point is not to get them to stop using AI – this is not possible – but to do so in a way that they can still learn and engage their critical thinking.

Redesign your assessments

There are two different types of assessment strategies as posited by the authors, one that eliminates the potential for generative AI use and the other that provides potential for collaboration with generative AI.

With the first, teachers can try in-class assessments, interactive oral assessments, supervised on-campus exams and tests as good ways to remove the risk of AI completely, but it is impossible to have every assessment be at school or at campus.

Which leads to the second type. Provide assessments in which students are told to “use AI to suggest ideas, summarise resources, and generate outlines/structures for assessments. They provide the AI completions as an appendix to their submission.”

In another possible assessment, students are instructed to “use AI-generated responses as part of their research and discovery process. They critically analyse the AI response against their other research. The AI completion and critique provided as part of the submission.”

Essentially, tell students to use AI as part of their assignments and give them instructions on how AI can help them with the process, such as research. Students will learn prompting, and teachers will be able to gauge learning because students must provide their AI work as part of the assignment.

“The lane 2 assessment might involve students collaborating with AI such as Bing Chat (which is internet-connected) to perform market research and competitor analysis, and other AI such as Adobe Firefly for the visual elements of campaign design,” they explain.

“Students document their interactions with the AI tools, including the AI’s initial market research and analysis and their critique and fact-checking processes to evaluate the AI’s outputs. Students also critique whether AI provided novel insights and whether it missed critical factors.”

“This is then presented live in class. The grading of the assessment is more heavily weighted on the documented process of critical co-creation.”

This type also decreases the possibility of cheating, because students must “show their work” and are already being allowed to use AI to make their lives easier. For the lecturers from Sydney, there is no third option.

“We do not foresee a viable middle ground between the two lanes. It needs to be assumed that any assessment outside lane 1 (i.e. that is un-secured) may (and likely will) involve the use of AI.”

That is a scary thought and one we agree with. There are two options here. Either do assessments in class, in person and away from AI software or assume that AI will be used and integrate this usage into assessments. This is the way things will go.

You can find the full guide to handling AI assessments here.