Orca Advisory

AI literacy · curriculum · learner agency

Shaping Agency: Designing a Curriculum for the AI-Literated Student

A practical framework for helping young people move from passive consumption of automated answers towards critical, ethical and self-directed use of artificial intelligence.

By Ken Corish · Orca Advisory · 10 min read

88% of young people aged 13–15 reported using AI tools for learning and creative activities.
96% of those aged 16–18 reported using AI tools multiple times each week.
72% of US teenagers were reported to have interacted with AI companion tools.

Recent research reveals a striking reality: 88% of teenagers aged 13 to 15 and 96% of those aged 16 to 18 are already using artificial intelligence tools for learning and creative activities multiple times a week. In the United States, 72% of teenagers have interacted with AI companion tools, with some even choosing to discuss deeply personal issues with chatbots rather than another human.

AI is no longer a future concept; it is the current terrain of youth culture.

The urgent question for educators and school leaders is no longer if students should use AI, but how we design a curriculum that transforms them from passive consumers of automated answers into self-directed agents of their own learning. If our curriculum merely treats AI as a technical skill, we miss the point. We must cultivate learner agency, critical thinking and social responsibility.

Here is how we can build a curriculum that empowers young people to command, rather than capitulate to, artificial intelligence.

Grounding Our Vision: Existing Global Frameworks

We do not have to build this curriculum from scratch. Pioneering international organisations have already mapped out what it means to be truly “AI-literate”.

EU & OECD AI Literacy Domains

A progression from recognising and using AI towards actively testing and improving the systems that shape people’s lives.

Engage

Identify AI, understand its influence, spot bias and verify outputs against reliable evidence.

Create

Use AI for brainstorming, design and production while retaining ownership, attribution and human intent.

Manage

Vet tools, judge when AI is appropriate and allocate work responsibly between humans and machines.

Shape AI

Test datasets and outputs, examine fairness and inclusivity, and propose improvements to system design.

The UNESCO Student AI Competency Framework establishes that teachers must act as guardians of safe and ethical practice in an AI-rich school environment. Building upon this, the European Commission and the OECD published their AI Literacy Framework: Empowering Learners for the Age of AI. This non-binding framework outlines 19 specific competences across four critical domains:

  1. Engaging with AI Learning to identify where AI is hidden in everyday products, understanding its influence and checking outputs carefully for accuracy and bias.
  2. Creating with AI Using AI during brainstorming, design and content production while maintaining ownership of original ideas and intellectual property.
  3. Managing AI Deciding when AI is—and is not—appropriate for a task, and knowing how to allocate work safely between a human and a machine.
  4. Shaping AI An advanced tier in which students evaluate training datasets, test outputs for fairness and suggest modifications to make systems more inclusive.

By adopting these domains, our curriculum shifts from teaching children how to operate a specific tool to cultivating the core attitudes of reflection, curiosity, adaptability and empathy.

Effective Use: Supporting Learning, Research and Content Creation

True student agency is cultivated when AI is used to deepen, rather than bypass, the learning process.

Overcoming “Cognitive Offloading”

Neuroscience tells us that deep, lasting learning requires “desirable difficulties”—cognitive tasks that demand genuine mental effort and struggle. When students retrieve facts from memory, work through complex mathematical proofs or write essays from scratch, their brains are actively building neural pathways.

The cognitive shortcut If a student asks AI to write homework, summarise a book they have not read or solve a physics problem for them, they may obtain a completed product while skipping the precise thinking processes required for learning.

The Flipped, Socratic Classroom

To counter this risk, our curriculum must shift pedagogy from the traditional “sage on the stage” towards a “guide on the side”. Under a Socratic-AI classroom model, the traditional sequence of learning is inverted:

The Socratic-AI Learning Sequence

AI may support initial access to knowledge; the classroom remains the place for challenge, dialogue and judgement.

Phase 1

Flipped Knowledge Acquisition

Students use approved, adaptive platforms for direct, initial instruction in foundational concepts.

Phase 2

Socratic Classroom Discussion

Teachers challenge assumptions, provoke analysis and guide collaborative problem-solving through purposeful questioning.

Separating the Work of Human and Machine

In practice, students must be taught how to manage the creation process actively. In a writing lesson, for example, a curriculum task could ask students to map the stages of producing an essay and decide explicitly which stages require human voice, reasoning and judgement, and which mechanical elements may be supported by AI.

Distinctively human work

  • Purpose, viewpoint and original ideas
  • Ethical judgement and contextual understanding
  • Interpretation, evaluation and final responsibility
  • Personal voice and creative intent

Potential AI-supported work

  • Generating alternative starting points
  • Reformatting or organising material
  • Exploring counterarguments for evaluation
  • Mechanical drafting under close supervision

When conducting research, students should never treat an AI answer as absolute truth. They must approach the output with scepticism and triangulate it against trusted primary and secondary sources before accepting it.

The Crucial Core: Ethical, Legal and Environmental Stewardship

An agency-driven curriculum must address explicitly the societal, legal and environmental implications of using these technologies. These lessons should be integrated across subjects—including Computing, Science, PSHE and History—rather than treated as a separate technical elective.

Intellectual Property and Data Privacy

Students must understand their legal rights and responsibilities. A student automatically owns the copyright in their original intellectual and creative work.

The marking trap Teachers and students should understand the implications of uploading essays, artwork or other original material into unapproved consumer AI tools for marking, feedback or summarisation. Such services may retain inputs or use them to improve public models.
  • If a tool is unapproved, it may use a student’s work to train or improve its public models.
  • Students should also be warned about secondary copyright risks when using and publishing generated content that may draw upon unlicensed source material.

The Environmental Price Tag

AI is resource-intensive, and a modern digital curriculum must highlight its ecological impact. Students should encounter the physical footprint behind apparently immaterial digital activity.

One generated AI image

The source material compares its environmental impact with charging a mobile phone three times.

A 30-second deepfake video

The source material compares its electricity use with streaming the complete television series Friends 500,000 times.

By making digital resource use visible, we teach students to ask whether AI is genuinely necessary for a task.

The goal is sustainable and responsible digital citizenship—not frictionless consumption.

Amplifying Voice and Increasing Agency

We cannot build a strategy for students without including them. True agency means involving young people in the governance of how technology is used in their schools.

Involving the Student Council

Schools that achieve the greatest success in AI integration can establish a Student Council or AI Youth Advisory Group to sit alongside school leaders. When reviewing policies on academic integrity and assessment, leaders should bring students into the room.

How do you feel about your work being evaluated or graded by an algorithm?

Do you feel comfortable being monitored by an automated system, or do you expect your teacher—who knows your social context and who you are—to remain the final judge of your progress?

By giving students a platform to debate these questions, we teach them that technology should serve human relationships, not automate them. They learn to challenge algorithmic bias, represent their community’s values and act as active digital citizens rather than passive users.

Conclusion: Empathy and Relationship Over Automation

As our classrooms become increasingly digital, the teacher’s role in supporting emotional, social and critical development becomes even more vital. AI possesses no authentic intent, understanding or empathy. It cannot replicate the rapport, understanding and trust that a great educator brings to a child’s life.

By designing a curriculum that integrates clear global standards, insists on critical thinking, enforces ethical and legal boundaries, and amplifies student voices, we do not merely prepare students for an AI-enhanced career. We prepare them to become independent, empathetic and highly capable human leaders who control the technology, rather than allowing the technology to control them.

Discuss your AI literacy curriculum