1. Protect Privacy. Without Exceptions
Student data is not a commodity and cannot be stored or exchanged freely.
How to comply:
- Limit data collection to what is strictly necessary. Collect only essential data required for the AI tool to operate. Minimize exposure of sensitive information.
- No profiling unless it directly supports learning needs. AI tools have the capacity to expand data on a user by monitoring and utilizing all interactions with the tool or platform. These can then be used on top of data collected directly from the user to build or infer user profiles. This should only be allowed when needed to monitor or enhance learning progress.
- No selling or sharing of student data or any data derived from it. Student data must remain confidential and not be monetized or shared with third parties. Cost of educational AI tools should be direct and visible instead of hidden costs in the form of data accumulation.
2. Fresh Start: No Sticky Labels
Children need room to evolve as they grow up. Social experimentation is critical while building one's character. Our cognitive and emotional capacities expand and we discover new tools to experiment with. I want my child to keep having an opportunity to go up on stage even if she hated being there back in 4th grade. Of course, we can "code" this in and make sure to challenge her every once in a while, but this applies to far subtler issues and is much different than starting afresh. AI can easily become an educational companion that builds a very detailed idea of who you are, but never forgets.
How to comply:
- Commit to giving students a fresh start every year. Reset the data collected on students at the beginning of each school year. Avoid AI-imposed labels for life.
- Allow students to reset and redefine their profiles. Provide opportunities for students to start afresh, reflecting their growth and development.
- Ensure AI-driven personalization does not limit learning potential. AI should adapt to students, but not box them into predefined pathways that restrict intellectual growth. Keep this in mind when selecting and properly configuring AI-based tools.
- Do not let past AI-based assessments limit future opportunities. Ensure that historical data does not hinder a student's potential or access to new opportunities.
3. Let Them Flourish: Use AI as a Tool, Not a Crutch
Early dependence on technology and AI carries the risk of hindering a child's development. Technology must be used to support, not disrupt, essential developmental milestones in young people.
Young people need to move. They need to physically engage in their world, carry out social interactions and get to explore their physical and emotional reactions to these interactions. They must build a sense of who they are and develop independent thinking and problem-solving abilities. They should improve their use of language, build a habit of reading, learn how to research and make connections across subject matters independently. All of these are at risk, if they become overdependent on technology and AI.
How to comply:
- Provide sufficient time away from technology. We know it sounds appealing to raise young people with a strong grasp of technology. However, we are responsible for leaving them the room to grow into capable, authentic individuals. They need to learn to balance and we need to support.
- Ensure AI complements, not replaces, human learning experiences. Integrate AI in ways that enhance other learning experiences including real-world projects and human interactions.
- Prioritize core human skills. Ensure that we are supporting and monitoring the development of a young person, prioritizing core human skills. This requires looking closely at the values of the learning environment, the goals that we set for the child and our approach to assessment. Start with exploring our guide on 21st Century Skills >>
- Encourage active learning over passive consumption. AI-driven learning must challenge students to think, experiment and create rather than just absorb AI-generated answers. This is tricky, but can be partially governed by selecting and properly configuring the right set of AI-based tools. It also requires challenging traditional modes of assessment that were based on presentation of information. See below "Principle 9: Mind-Shift" to learn more on this.
4. Free from AI Bias
AI typically reflects the biases of its creators and the data it was trained on. Unchecked, these biases can reinforce stereotypes and unacknowledged myths. Young people should have the right and the space to form their characters, opinions and attitudes.
How to comply:
- AI tools must undergo rigorous bias testing. Schools must demand transparency from AI providers, ensuring models do not reinforce common biases. AI providers should also ensure that the models are trained on sufficiently diverse datasets. It will be impossible to audit AI against all forms of bias, but a proper audit can reveal several common issues. Read more on this on "Principle 7: Transparency" and "Principle 8: Accountability."
- Students should learn to detect and challenge AI biases. AI literacy must include critical analysis skills and a common understanding of AI flaws, so that students recognize and push back against flawed thinking patterns. See more on "Principle 5: AI Literacy" below.
- Educators and students should have the power to override AI-driven suggestions. AI should advise, not dictate. Educators and students must have the final say in determining students' learning paths.
5. Access: Avoid Building Invisible Walls
AI should ideally expand access to learning, not create further digital divides. If not carefully managed, AI can reinforce social or educational inequalities. Not having access to the best AI tools is one potential risk that threatens students from lower socioeconomic backgrounds. The other risk is for AI to be utilized as an admissions tool, serving as a gatekeeper for higher education programs or opportunities.
How to comply:
- No paywalls for essential AI learning tools. Schools must ensure that AI-powered educational tools are accessible to all students, regardless of socioeconomic background.
- AI should adapt to different learning environments. AI systems must be designed to function across varied school settings, including low-resource communities. This includes:
- Operating with unreliable Internet connectivity
- Operating on shared devices
- Supporting multiple languages
- Ensuring compatibility with older or low-cost devices
- Using diverse datasets to minimize biases against different cultural backgrounds and learning styles
- Allow for adjusting content to align with local curricula and needs
- Ensure transparency in AI-driven admission processes. Hidden biases of AI models must not create hidden barriers to scholarships, advanced courses or university admissions. This connects to "Principle 7: Transparency" and "Principle 8: Accountability."
- Do not penalize students based on previous exposure to AI-based learning. Ensure that no student is disadvantaged by AI-driven assessment systems due to lack of prior exposure to AI-based learning. Design systems and processes to allow for inexperienced AI users to navigate easily.
6. AI Literacy: Teach AI Awareness
Every student must have a solid understanding of how AI works, where it fails (potential flaws) and how it influences decision-making. AI literacy is as essential as reading, writing and math in the 21st century.
How to comply:
- Students must be taught AI fundamentals. Include AI education in curricula, covering topics like data bias, decision-making algorithms, information mistakes, privacy risks, ethical risks and even prompt writing techniques.
- AI-generated content must be questioned, not blindly accepted. Teach students to critically evaluate AI outputs, cross-checking information rather than assuming AI is infallible.
- Encourage hands-on AI experimentation. Let students build and test AI models to understand their strengths and limitations firsthand.
- Full transparency on how AI influences their learning. Educate students about AI's role in their education as well as possible uses of AI in assessments or as part of future admissions into advanced programs.
7. Transparency: No AI in the Shadows
AI must not operate behind closed doors. Every school using AI must disclose how and why AI is used, what data it collects and how it affects students. Educators and students should also be informed about where it is OK to use AI and how it needs to be disclosed.
How to comply:
- Schools must publish clear AI policies. No hidden algorithms. Students, educators and parents deserve full transparency on how AI shapes learning as well as where and how AI should be used. Get in touch with us here and we will be happy to send you a Sample AI Guidance Document that you can adapt to your institution's needs.
- AI use in assessments and grading must be disclosed. Students have the right to know if AI influences their admission to programs, grades, learning paths or academic progress.
- AI decisions must be explainable. Schools must ensure that any AI-driven recommendations, whether in learning paths or discipline enforcement, can be reviewed and challenged. In that sense, AI outputs should serve as an input into the final decision and AI should not be the final decision-maker.
8. Accountability: Hold AI Builders Accountable
The responsibility for ethical AI does not rest solely on educational institutions. AI models are very complicated and get trained using very large datasets. Even their own developers do not fully understand the details of the decisions made by AI algorithms. It is impossible for each educational institution to individually question and audit each AI tool.
Instead, schools and parents should demand accountability from tech providers. Companies building AI tools must be held to rigorous safety, fairness and transparency standards.
How to comply:
- AI developers must submit to periodic independent audits. Schools should only adopt AI tools that have undergone thorough external evaluations for algorithmic fairness and security. Ensuring safe and ethical AI should be the principal goal.
- Tech companies must disclose training data and methodologies. AI models should not be black boxes.
- Ensure clear channels for reporting AI failures. Educational institutions must have mechanisms for their educators, students and parents to flag AI errors, biases or harmful outputs.
- No AI tool should be considered "final" - continuous review is essential. AI systems must be regularly updated to align with evolving educational standards, safety regulations and ethical concerns.
9. Mind-Shift: Rethink Education in the AI Era
AI has the potential to radically transform education for the better - but only if we are willing to let go of outdated models. Traditional education, built around the delivery of information and the assessment of its retention, is fundamentally incompatible with an AI-powered world. And that’s a good thing.
For decades, educators have called for a shift away from the traditional mode of teaching, but adoption of such change has been extremely slow. Now, AI has made information instantly accessible and universally available. This levels the playing field and forces all of us - educators, administrators, institutions - to confront a hard truth: information delivery is no longer the primary value we bring to learners. The future of education must be about meaning-making, creativity, collaboration - getting to know oneself and developing as a human. See more on these critical skills by reading our guide on Future Skills >>
This change is not just about adding AI-based tools to existing systems. It demands a paradigm shift - a personal and institutional transformation in how we think about teaching and learning. We owe it to the young people we serve to evolve with the times. The outcomes will be worth the effort.
How to comply:
- Engage in deep self-reflection and professional assessment. Assess your current teaching approach and what it means to be a teacher or a learner in the new world. Identify areas for change and growth in light of this new paradigm.
- Commit to professional transformation. Participate in training programs designed to help educators rethink their roles, shift their mindsets and develop practices aligned with 21st-century learning.
- Transform the whole institution, not just individuals. Partner with LearnButWhy to assess your school’s operations and learning climate. We help guide educational institutions through a collective transformation toward learner-centered values and AI-aligned practices.
- Build or join a professional learning community. Connect with other educators committed to change. Share insights, challenges and get support as you reshape your practice.
- Adopt the right AI and edtech tools intentionally. Choosing tools is not just about features - it's also about learner needs and safety standards. LearnButWhy helps institutions curate and implement a learner-centered AI stack that aligns with their learning goals and educational mission.
