Sandy RobinsonWe don't need more AI tools.
We need better questions.
Permission to Learn.
Moving from fear to inquiry in AI classrooms.
A year of embedded observation in secondary classrooms. Argues that a teacher's emotional stance toward AI shapes whether students treat it as a thinking partner or a shortcut. Introduces VERIFY and EPFL as frameworks for moving from banning to teaching. References Hasan (2024), Porayska-Pomsta et al. (2024), and UNESCO (2021).
Classrooms reveal their mood before anyone speaks.
Many students have learned to treat curiosity like it carries a risk.
Misuse stems from disconnection, not deception.
Teaching in the age of AI is often defined by control. It does not have to be.
Latest posts.
Auto-pulled from @MrsSandyRobinson on Medium. New posts appear here automatically.
Loading recent posts…
Where to start.
A few key pieces rather than the full archive. Some are early versions of ideas that have since developed. The complete list is on LinkedIn.
VERIFY: A Framework for Teaching Students to Think With AI
The first full write-up of VERIFY as a six-step thinking cycle. Argues the bigger risk isn't misuse but students handing over their thinking. Maps onto the Australian Framework for Generative AI in Schools. The version on the Work page has been updated since.
Read on LinkedInThe Ethical Prompting Feedback Loop
The original six-step EPFL: Feeling, Intent, Prompt, Output, Consequence, Reflection. Argues that the emotional and ethical side of every AI interaction is the part we usually miss. Version 2.0 on the Work page came from using v1 in real classrooms.
Read on LinkedInBeyond the Buzzword: What 'Ethics' Really Means for AI in Education
The piece where VERIFY first appeared (before becoming a standalone article). Argues that "ethics" gets used as a reason to pause progress instead of shape it. Ethics aren't delivered through policies; they show up in how you act every day.
Read on LinkedInDon't Start With the Chatbot
Argues we shouldn't start AI education with chatbots but with thinking. Challenges the cheating frame, makes the case for play-based learning that builds critical thinking before tool access. Introduces RePrompted, the card-based classroom tool.
Read on LinkedInThe Ethics of Literacy in an AI World
Argues that treating literacy as reading and writing is not just outdated but ethically harmful. Meaning is already multimodal. Connects multimodal literacy to AI ethics: ethical agency begins with recognising how meaning is designed across text, image, sound, data, and interface.
Read on LinkedInData Power: The Missing Piece in Digital Literacy
Challenges the myth that technology is "just a tool." Argues students need to understand how their digital footprints are shaped, how data flows invisibly, and how AI systems influence their worldview. From digital users to digital shapers.
Read on LinkedInThis Is How We Build Better Thinkers
Observes how students freeze when research hits obstacles. Paywalled articles, broken links, dead ends. Argues the key is teaching students to shift, not stop: use AI as a sounding board, work backward through reference lists, make research their own.
Read on LinkedInReader responses.
This is exactly the adoption mistake many institutions make: they treat access as literacy. Giving students a chatbot is not the same as teaching them how systems frame, distort, compress, and reward certain kinds of thinking.
Ethical AI use in education isn't just policy, it's daily practice. The conversation about ethics needs to happen before the tech is even turned on.
Follow the writing
as it develops.
New essays on Medium, cornerstone pieces on LinkedIn.