91��Ƭ��

Stephanie Holt
Director of Teaching and Learning, DSB International 91��Ƭ��, India
Co-author of AI for Learning: 101 Assessments K-12 Unlocking Mastery of AI (Belgravia Press, 2024)
Alexander Harris
Deputy Head Academic, Sanford International 91��Ƭ��, Ethiopia

In the buzz and bustle of AI in education, it’s tempting to get caught up in the race to acquire the flashiest edtech. Some schools proudly announce the adoption of AI platforms, the integration of chatbot tutors, or the piloting of algorithmic grading tools, while others bury their heads in the sand and hop ethat it will all pass them by. But as we stand on the cusp of a seismic shift in how students engage with knowledge, a more important truth emerges: AI in education is not about what technology you adopt—it’s about how your students learn to use it.

The focus must shift from learning to use AI to using AI for learning. That distinction isn’t just semantic—it is pedagogical, ethical, and profoundly transformational.

From Tool to Tutor: Repositioning AI in the Learning Process

At its best, AI should function like a compass, not a crutch. When students use generative AI responsibly—whether for drafting essays, analysing historical trends, or solving mathematical problems—they’re not outsourcing thinking. They’re enhancing it.

The key lies in guided, purposeful use. Research cited in AI for Learning: 101 Assessment Strategies for K-12 91��Ƭ��s Unlocking Mastery of AI reveals that students given structured support in AI use significantly outperform peers who use AI unguided or uncritically (Holt & Harris, 2024). Left to its own devices, AI becomes a shortcut. Integrated into a structured learning journey, it becomes a scaffold for deeper inquiry.

The Traffic Light System: Teaching Ethical and Effective Use

One practical approach to this balance is the Traffic Light System. In this model, AI use in student work is categorised into three bands:

• Green – Enhances learning without replacing it (e.g., brainstorming prompts, language practice, data analysis).
• Amber – Risks displacing the learning process and needs monitoring (e.g., using AI to paraphrase or summarise without reflection).
• Red – Replaces learning or violates academic integrity (e.g., full AI-generated assignments).

This framework, introduced in AI for Learning, provides clarity for students and teachers alike, distinguishing between collaboration and collusion. It helps learners understand when AI acts as a learning partner—and when it undermines their growth.

AI is Pedagogy, Not Product

We need to think of AI not as a plug-in but as part of a pedagogy of possibility. As AI for Learning outlines, the most meaningful AI integration isn’t platform-dependent. It happens when:

• Teachers model metacognitive reflection—asking not just “What did you write?” but “How did AI shape your thinking?”
• Students are encouraged to interrogate AI outputs, question accuracy, and add their own insights—developing digital and critical literacy simultaneously.
• Learning objectives drive AI use, not the other way around.

This is the heart of the book’s Assessment-as-Learning principle: assessment isn’t just about measuring what students know. It’s about helping them learn through the process of inquiry, revision, and reflection—with AI as a thinking partner, not a ghostwriter.

Equity and Access: Avoiding a Two-Tier AI System

Another urgent issue the book raises is equity. If AI tools are only accessible to students whose families can afford subscriptions or devices, we risk deepening existing achievement gaps. AI for Learning recommends school-led solutions: centralised subscriptions, in-school access, and parent workshops to ensure that AI becomes a bridge, not a barrier.

Education leaders must ask: Are we equipping all students—not just the privileged few—to harness AI responsibly and reflectively?

Assessing the Learner, Not the Output

One of the most provocative questions AI for Learning poses is this: “What are we really assessing when students use AI?”

Traditional assessment often focuses on the final product. But when AI is in the mix, educators need to shift the lens. Did the student understand how to refine the AI output? Did they cross-check the facts? Did they integrate their voice, their argument, their evidence?

To support this shift, the book offers over 100 assessment strategies categorised across five learner levels—from foundational to mastery. These assessments aren’t about checking AI usage; they’re about checking learning through AI usage.

Cultivating AI Fluency, Not Just Compliance

Ultimately, schools need to prepare students not just to use AI but to thrive in a world shaped by it. That requires more than teaching prompt engineering or banning ChatGPT. It demands a fundamental change in how we define success in school.

Instead of asking “Can students write this without AI?”, we must begin asking:
“Can students show us what they’ve learned with AI?”
“Can they reflect on how they got there?”
“Can they distinguish AI’s voice from their own—and use that difference to grow?”

That’s what fluency looks like. And it starts not with the tech we buy, but with the trust we build—in our learners’ ability to engage, critique, create, and reflect.

It’s Not the Tool—It’s the Thinking

91��Ƭ��s that treat AI as an add-on will never unlock its power. 91��Ƭ��s that treat AI as a thinking partner, a means to deepen learning, will transform education. Not because of the tools—but because of the thinkers they help shape.

As AI for Learning reminds us: “Used well, AI becomes the Aristotle to a learner’s Alexander. Used poorly, it is as poor a learning tool as ‘cut and paste’” (Holt & Harris, 2024, p. 41).

Let’s teach our students to lead—not follow—their tools.

Reference
Holt, S. & Harris, A. (2024) AI for Learning: 101 Assessment Strategies for K-12 91��Ƭ��s Unlocking Mastery of AI. London: Belgravia Press.

About the authors

Stephanie Holt is an educator with over 20 years of experience, having worked globally in various capacities including as an Advanced Skills Teacher of English in the UK, 91��Ƭ�� Improvement Officer, Vice-Principal in Malaysia, and Deputy Head in Moscow. Currently, she is the Director of Learning and Teaching in Mumbai.

Involved with the OECD Classrooms+ initiative, Stephanie has delivered workshops for COBIS on metacognition and using AI for Learning, will be speaking at the OECD Classrooms+ conference 2025 and was a keynote speaker at the WCE Conference 2024. Her forward-thinking approach has been recognised by her shortlisting for the GESS Award 2024 for Positive Change in Education.

She has co-authored the book “AI for Learning: 101 Assessment Strategies for K-12 91��Ƭ��s” with Alexander Harris. Stephanie is a thought leader in AI and education, contributing regularly to global conversations on enhancing learning outcomes through innovation. Her research as a PhD candidate for Brunel University London and work at DSB International 91��Ƭ��, Mumbai significantly enhances educational practices, empowering educators and fostering student success.

Alexander Harris is an educational leader with a global career spanning multiple continents. He holds a Master’s in Educational Leadership with Distinction from UCL’s Institute of Education and has held senior roles in prestigious international schools. Alexander is known for his innovative approach to AI integration and curriculum design, driving academic growth and fostering ethical leadership.

His expertise in change leadership has transformed educational communities, empowering educators to create dynamic, student-centered learning environments. Alexander has led successful AI-driven initiatives that enhanced student engagement and achievement.

He is the author of two upcoming books on education and numerous novels and plays under the pen name ‘Thomas Alexander.’ As a sought-after speaker, Alexander shares visionary insights on AI in education, curriculum design, and leadership development. His work is grounded in servant leadership, promoting integrity and equity as transformative forces for good in education.