With AI-tools like ChatGPT increasingly available to students, educators worldwide are working to re-imagine both assignments and tests as a way to preserve academic integrity while also promoting genuine learning. Here’s how teachers are adjusting and why it matters.
Reimagining Assignments to Combat AI Abuse
In British Columbia, teacher Gary Ward and his colleagues have started personalizing homework by including prompts that reflect a local situation or a recent specific classroom discussion to make it so individualized as to be almost impossible for AI to copy. They’ve even tested tools that can analyze how likely a prompt is to be completed by AI, which helps them design more “cheat-resistant” tasks.
Back to Analog: Writing & Orals
Some educators are embracing old-school formats. The handwritten paper, the in-class essay or exam and the oral quiz are all methods some instructors use to test whether a student has really composed his/her own work. Richard Griffin in Manchester moved major assignments to discussions and responses (and to personal responses) — it had been a traditional paper-based approach — while Paul Shockley at StephenF.Austin State University has abandoned AI tools entirely after he encountered students turning in papers generated by AIs. He is now interested in experiential and reflective work, which requires original thought.
Valorizing Thinking Over Convenience
For Professor Shockley and others, the aim is not to outlaw A.I. altogether (A.I. can help inspire ideas) but that it doesn’t replace deep thinking. Asking students to relate their coursework to personal experience or current events gets them personally engaged and thinking deeper, teachers say.
A Larger Change in Pedagogy
Colleges are not the only ones experimenting with a new mode of teaching.
This shift goes well beyond individual classes:
UK universities revealed some 7,000 verified cases of AI-related cheating in 2023–24, leading to a re-evaluation of testing procedures.
Stefania Druga, who researches at Google DeepMind, is promoting co-creative learning, in which students use AI to program games or other projects instead of write monster English essays so they can learn “how things work out in the world”.
Reid Hoffman (cofounder of LinkedIn) sees exams taking the form of AI “examiners”, for oral exams to be used to verify understanding.
What This Means for Students & Schools
Trend:
- Assessment shifts
- Assignment redesign
- AI integration
- Academic vigilance
Impact:
- 2020 has seen everything from essays to hand-written assignments, oral exams and class debates
- “These projects include a personal or contemporary context that AI is unable to mimic right now,” it continues.
- Teaching AI as co-creation, not cut and paste
- Detection tools are not enough; pedagogy is the operative word
We understand.” We toe a line precariously juxtaposing the benefits of AI with preservation of critical thinking and ethical scholarship. Instead, we’re seeing a healthier approach: AI that improves how students learn, while also deliberately creating tasks that demand real student insight, creativity and reflection.
This shift in education demonstrates resilience and determination to keep true learning at the heart of schooling, despite the influence from potent AI tools.
