AI in the service of education - Practical guide for teachers and educational teams

This guide on AI and education first shows that learning to use AI in schools, with an inclusive pedagogy, prepares future adults to navigate a digital environment that directly impacts their health (stress, screen use, exposure to misinformation). It emphasises digital literacy and the development of critical thinking: these skills are essential for identifying false health information, understanding risks, and making informed choices. The CUA logic and AI-supported differentiation help to reduce learning inequalities, which are themselves major determinants of long-term health. Working on the classroom climate, cooperation, and reasoned debate strengthens key psychosocial skills for health promotion (self-expression, emotional regulation, conflict resolution). Finally, the ethical approach and co-construction of rules for the use of AI in schools can inspire similar initiatives.in health prevention, where transparency, participation, and equity are central.

Source: 📒 AI in the Service of Education - Practical Guide for Teachers and Educational Teams

  📜🔗LINK 

Number of pages: 48

1. Analytical summary

1.1. Context, audiences, and issues

The guide is aimed at teachers and educational teams who are encountering generative AI in the classroom without always having clear pedagogical, ethical, and regulatory benchmarks. It emphasises the necessity of "talking pedagogy above all else": technology is merely a tool in service of learning objectives, differentiation, and inclusion, particularly through frameworks such as Bloom's taxonomy, Universal Design for Learning (UDL), or institutional pedagogy (part 1, p.4–8). The work demonstrates how AI can personalise situations (problems contextualised to students' interests, narratives where the student is the hero) while also reminding of the risks: loss of meaning if one relies solely on the "magic" of the tool, issues of equity, pedagogical quality, and academic integrity (parts 1 and 4, p.3, p.43–45).

1.2. Operational contributions for field actors

The document provides a range of concrete practices: ready-to-use prompts, activity ideas, examples of tools (MagicSchool, History Timelines, Debate AI, AutoClassmate), differentiation matrices, and CUA or Bloom-oriented query formulations (parts 2 and 3, pp. 12–18, pp. 21–23, pp. 33–42). It equips teachers to generate multiple-choice questions, timelines, personalised narratives, multimedia resources, differentiated assessments, lesson plans, and sequences, systematically linking these uses to existing pedagogical frameworks. It also offers a series of prompts for professional development (classroom management, climate, inclusion, project-based pedagogy), making AI a "first mirror" before potentially engaging in discussions with peers or trainers (part 3, pp. 29–32). Finally, the section on "perspectives and regulation" refers to international resources (TeachAI, Office of Educational Technology, institutional policies) to co-construct usage charters and local guidelines.

2. Key points of the document

1. AI as a pedagogical assistant, not as a pilot (pp. 3–7)

   The guide clearly states that AI is an assistant serving the pedagogical compass: it is the teacher who defines competencies, tasks, assessment modalities, and framework choices (Bloom, CUA, institutional pedagogy, cooperation, memorisation, projects). AI is used to generate content, resources, or scenarios tailored to different cognitive levels, learning styles, and specific needs (pp. 5–7).

2. Personalisation of learning based on students' interests (pp. 8–10)

   The work illustrates how AI allows for the creation of problems, texts, and narratives contextualised to students' interests (sports, fashion, dinosaurs, etc.), with observed effects on motivation, engagement time, and the quality of social interactions in the classroom (p.8–10). The teacher remains in control of the objectives and progression, with AI serving to quickly multiply the variants of equivalent situations in terms of the targeted skills.

3. AI and Universal Design for Learning (UDL) for inclusion (p.12–16)

   An important section shows how AI can accelerate the implementation of UDL by generating varied modalities of representation, action/expression, and engagement (read-aloud texts, visual supports, interactive games, differentiated pathways). The guide offers prompts for ChatGPT or dedicated tools (MagicSchool) to construct UDL tables, complete sequences, and adaptations (accessibility, assistive technologies, simplified content, audio formats, etc.) (p.12–16).

4. Systematic pedagogical differentiation using typed prompts (p.33–38)

   Part 3 structures differentiation around four levers: content, process, products, and learning environment, with lists of prompts to personalise, simplify, enrich, diversify activities, adjust pace, or vary outputs (written, audio, video, collaborative projects) (p.33–37). It shows how AI can support remediation, enrichment, self-assessment, and co-construction of pathways, without lowering skill requirements (p.34–38).

5. Towards informed and co-constructed regulation of AI in schools (p.43–45)

   The last section addresses ethical, regulatory, and governance issues: academic integrity, plagiarism, bias, data respect, clarity of usage rules for students and teams (p.43–45). It refers to resources such as TeachAI (Code.org, ISTE, ETS, Khan Academy) and guides from educational authorities to develop local charters for good practice, co-constructed with students, teachers, parents, and management.

3. Action points for local stakeholders

1. Integrate AI into school and educational territory projects (p.3–7, p.29–33, p.43–45)

   Teams can include an "AI and pedagogy" project in the establishment or priority education network project, relying on the proposed frameworks: Bloom, CUA, differentiation, cooperative pedagogy (p.5–7, p.33–37). An internal training plan can combine workshops for co-writing prompts, analysis of scenarios from the guide, and collective work on a usage charter inspired by the cited resources (TeachAI, institutional guides, p.43–45).

2. Co-construct with teams a "local bank" of prompts and scenarios (p.21–25, p.33–38)

   The guide proposes a "collaborative prompt bank" approach that teams can adapt at the establishment or community level (p.21–25). It is possible to pool, by discipline and level, tested and commented prompts (sequences, assessments, remediation, projects, debates, etc.), with work to adjust to national standards and field constraints (numbers, time, equipment) (p.25, p.33–38).

3. Use CUA and AI to strengthen inclusion and combat learning inequalities (p.12–16, p.33–37)

   Inclusive coordinators, AESH, special education teachers, and medical-social teams working in schools can rely on CUA examples and adaptation prompts (simplified content, audio formats, visual supports, differentiated activities) to better respond to the heterogeneity of students (p.12–16, p.34–36). It is necessary to add local validation procedures (human proofreading, quality control of educational content, bias checking) and clear guidelines for use with students in situations of disability or social vulnerability.

4. Equipping continuous professional development through AI as a 'mirror' of practice (p.29–33)

   Academic trainers, training managers in INSPE, complementary school associations, and school health networks can use the series of professional reflection prompts (class management, climate, non-violent communication, inclusion, differentiation) as supports for practice analysis (p.29–32). AI serves here as a first step of questioning, before a group perspective work (workshops, communities of practice) and confrontation with scientific and regulatory references.

5. Building a critical culture of AI and digital ethics with students (p.25–27, p.43–45)

   Teachers, CPE, mediators, and educators in citizenship education can use the proposed examples of debates, controversies, and timelines (History Timelines, Debate AI) to work on critical thinking and the analysis of AI outputs (bias, reliability, sources) (p.16–21, p.25–27). In parallel, they can co-develop with students an academic usage charter for AI (what is allowed/forbidden, conditions for assignments, transparency of uses), relying on the mentioned governance resources (p.43–45).

4. Additional References

🔍➕ For more information, see the articles referenced by "Pratiques en Santé" on the topic of AI and digital technology ➡️🔗https://pratiquesensante.odoo.com/2-6-intelligence-artificielle-numerique

1. Hyatt S.E., Owenz M.B. (2024), "Using Universal Design for Learning and Artificial Intelligence to Support Students with Disabilities" –College Teaching.

   Scientific article showing how to articulate AI and UDL to support students with disabilities, with varied assessment options and attention to digital literacy.https://discovery.fiu.edu/display/pub303433

2. OECD (2023), "Emerging governance of generative AI in education", chapter of theOECD Digital Education Outlook 2023. https://www.oecd.org/en/publications/oecd-digital-education-outlook-2023_c74f03de-en/full-report/emerging-governance-of-generative-ai-in-education_3cbd6269.html

   Comparative analysis of national policies on generative AI in education, including recommendations for adaptive frameworks, guidelines, and consideration of biases, equity, and data protection.

3. TeachAI – "AI Guidance for Schools Toolkit" (Code.org, ISTE, ETS, Khan Academy, 2024).

   International toolkit dedicated to educational leaders for building local AI policies in schools: concrete examples, charter templates, usage scenarios, and ethical guidelines.https://www.teachai.org/cs

5. Cross-sectional analysis — values of Pratiques en Santé

• Literacy: the guide adopts accessible language for non-specialist teachers in digital technology and clarifies the frameworks (Bloom, UDL) through concrete examples and ready-to-use prompts, without excessive technical jargon (p.5–7, p.12–16, p.21–23).

• Empowerment: students are positioned as active participants in learning through content based on their interests, timelines they create themselves, debates, collaborative projects, and by choosing modes of expression (p.8–10, p.16–21, p.25–27, p.33–37).

• Participation: the document describes mechanisms for co-constructing outputs (projects, debates, group work) and suggests student involvement in developing the rules for using AI (p.16–21, p.25–27, p.43–45).

• Community health: although the health theme is not central, the collective dimension appears in the focus on inclusion, classroom climate, cooperation, and responsible digital citizenship (p.3, p.12–13, p.25–27, p.29–31).

• Ethics: the issues surrounding AI are addressed in terms of risks, responsibility, codes of conduct, and limitations (plagiarism, bias, substitution for human expertise), with resources to formalise ethical rules (p.43–45).

• Human rights: the guide aims for accessibility and equity, particularly through the CUA, differentiation, and consideration of students in difficulty or with disabilities, without stigma (p.12–16, p.33–37).

• Intersectorality: links are outlined with academic research, international networks (TeachAI, Code.org, ISTE, ETS, Khan Academy), and educational authorities, but the 'social/medico-social' dimension remains underdeveloped (p.21–23, p.33, p.43–45).

• Partnership: models of collaboration between schools, universities, and specialised organisations are cited (Stanford, Office of Educational Technology, associations, tool publishers), suggesting alliances for training and regulation (p.32–33, p.43–45).

• Combating discrimination: the dimension of inclusion, diversity of profiles, and consideration of specific needs is explicit, but discrimination (gender, origin, disability, etc.) is primarily addressed through the example of the history of women's rights and the potential biases of AI, without in-depth analysis of systemic discrimination (p.18–21, p.43–45).

6. Assessment of the reliability of the resource

• Scientific relevance

  The guide is not a review of scientific literature but a practical tool based on recognised theoretical frameworks (Bloom's taxonomy, Universal Design for Learning, cooperative pedagogy, institutional pedagogy). It refers to external academic and institutional resources (UDL & AI for students with disabilities, AI governance guides, TeachAI tools), which enhance the credibility of the recommendations. The methodology is based on classroom scenarios, qualitative experiments, and examples of prompts rather than quantitative data; this is suitable for a practical guide but requires maintaining a critical perspective on the generalisation of the described effects.

• Operational relevance

  The resource is strongly action-oriented, with fine granularity (prompts, examples of activities, formulation suggestions, sequence ideas, named tools) directly transposable into various school contexts (primary, secondary, high school). It includes warnings about the limitations of the tools and emphasises the need to retain teaching expertise, contextualise, and co-construct usage rules, which enhances its relevance for teams seeking to frame AI within an inclusive and ethical education perspective.

7. MCQ – 5 questions

Part 1 — MCQ without answers

Question 1 (p.5–7)

According to the guide, what main role should AI play in education?

a) Replace the teacher in lesson preparation

b) Serve as an assistant to achieve educational objectives defined by the teacher

c) Fully automate student assessment

d) Impose standardised teaching methods

Question 2 (p.12–16)

In Universal Design for Learning (UDL), which dimension is NOT explicitly mentioned among the three structuring axes?

a) Representation

b) Action / expression

c) Engagement

d) Career guidance for students

Question 3 (p.8–10)

What effect was observed when using problem statements based on students' interests?

a) A decrease in motivation and engagement time

b) An increase in requests for disciplinary sanctions

c) A positive expectation of the slot and increased activity time

d) A disappearance of peer interactions

Question 4 (p.33–37)

Among the following proposals, which is NOT one of the four differentiation levers described for using generative AI?

a) Content

b) Process

c) Productions

d) Administrative service notes

Question 5 (p.43–45)

What is a central issue of the 'informed regulation' of AI in schools, as presented in the guide?

a) Totally prohibit AI in all learning contexts

b) Develop clear guidelines co-constructed with stakeholders

c) Entrust regulation exclusively to private publishers of AI tools

d) Replace school curricula with AI-generated content

Order of chosen correct answers: b / d / c / d / b.

Part 2 — Commented correction

Question 1 (p.5–7)

According to the guide, what main role should AI play in education?

✅ Correct answer: b) To serve as an assistant in the service of pedagogical objectives defined by the teacher

📝 Explanation: The guide emphasises that "the compass is indeed pedagogical": the teacher defines skills, objectives, and approaches, with AI being merely an assistant to generate content and resources (p.5–7, p.11). Source: p.5–7.

Question 2 (p.12–16)

In Universal Design for Learning (UDL), which dimension is NOT explicitly mentioned among the three structuring axes?

✅ Correct answer: d) Career guidance for students

📝 Explanation: UDL is described around three axes: modes of representation, action/expression, and engagement; career guidance is not listed as a foundational axis in this framework (p.12–13). Source: p.12–13.

Question 3 (p.8–10)

What effect was observed when using problem statements based on students' interests?

✅ Correct answer: c) A positive expectation of the niche and a greater amount of activity time

📝 Explanation: The example of the weekly problem-solving session indicates that students eagerly anticipate the slot, with increased motivation and activity time, and enriched social interactions (p.8–9). Source: p.8–9.

Question 4 (p.33–37)

Among the following proposals, which is NOT one of the four differentiation levers described for using generative AI?

✅ Correct answer: d) Administrative service notes

📝 Explanation: The guide structures differentiation around four levers: content, process, products, and learning environment; administrative service notes are not included (p.33–37). Source: p.33–37.

Question 5 (p.43–45)

What is a central issue of the 'informed regulation' of AI in schools, as presented in the guide?

✅ Correct answer: b) Develop clear guidelines co-constructed with stakeholders

📝 Explanation: Part 4 highlights the need for open discussions between teachers, students, parents, and management, as well as explicit guidelines based on resources like TeachAI and institutional guides (p.43–45). Source: p.43–45.

8. Frequently Asked Questions (FAQ)

1. Who is this guide aimed at and in what contexts can it be used?

   The guide is designed for primary and secondary school teachers, educational teams, management, and trainers, but it can be adapted to other educational contexts (academic support, adult training). It targets situations where generative AI is available or being introduced, particularly for lesson preparation and personalising learning (p.3, p.29). Pages: p.3, p.29–31.

2. How can one practically start with AI without being a digital expert?

   The guide recommends starting from educational needs, then using simple and contextualised starting prompts (level, subject, objectives) before gradually refining them (p.5–7, p.21–23). Examples of prompts for "beginners, intermediate, advanced" are provided for constructing sequences, activities, and assessment materials (p.21–23, p.33–35). Pages: p.5–7, p.21–23, p.33–35.

3. How can AI help differentiate learning in a heterogeneous classroom?

   AI can quickly generate variations of content (problems, texts, visual aids), adapt the level of language, propose remediation or enrichment activities, and provide diverse production options (written, oral, multimedia) (p.33–37). The guide organises these uses around four levers: content, processes, productions, and learning environment (p.33–37). Pages: p.33–37.

4. What are the specific contributions of AI for the inclusion of students with difficulties or disabilities?

   AI allows for the production of simplified texts, audio versions, visual materials, interactive activities, and personalised learning plans, in line with the CUA (pp. 12–16, pp. 34–36). External resources on AI and accessibility are mentioned, as well as a MagicSchool module for personalised plans (IEP). Pages: pp. 12–16, pp. 34–36.

5. How can AI be used for the professional development of teachers?

   The guide offers a series of reflective prompts on classroom climate, non-punitive management, non-violent communication, motivation, and differentiation, to help teachers analyse and enrich their practices (pp. 29–32). It also mentions research work (Stanford project 'Empowering Teacher') that uses AI to analyse class transcripts and formulate feedback (pp. 32–33). Pages: pp. 29–33.

6. What ethical and regulatory precautions should be taken before generalising the use of AI in schools?

   The guide emphasises the need to clarify usage rules (plagiarism, academic integrity, data confidentiality), to inform students and families, and to rely on guides and toolkits (TeachAI, institutional policies) (pp. 43–45). It suggests engaging in an open dialogue between teachers, management, students, and parents to co-construct appropriate rules (pp. 43–45). Pages: pp. 43–45.

7. Does AI replace teachers in lesson preparation and assessment?

   The guide is explicit: AI is just a tool that automates certain tasks (generation of exercises, materials, lesson plans), but it does not replace pedagogical judgement or the detailed knowledge of students (p.5–7, p.38–40, p.46). The teacher remains responsible for choices, verification, adaptation of content, and the construction of the ethical framework (p.11, p.43–45). Pages: p.5–7, p.11, p.38–40, p.43–46.

9. Rewriting in Easy to Read and Understand (FALC)

9.1. Analytical summary in FALC

Title: Why this guide?

This guide talks about artificial intelligence, also known as AI.

It helps teachers and school teams use AI to improve their teaching.

AI should remain a tool: it is the teacher who decides the objectives for the students.

The guide shows how to adapt lessons to each student with AI.

It also discusses justice, inclusion, and respect for rules in school.

Title: What the guide offers

The guide provides specific examples for using AI in the classroom.

It includes ideas for exercises, questions, and projects to do with students.

It offers ready-made phrases, called 'prompts', to give to the AI.

It helps prepare lessons, assessments, and activities for different levels.

It explains how to discuss the risks of AI and create rules with students.

9.2. Key points in FALC

Point 1: AI is a tool

AI does not replace the teacher.

The teacher retains control over choosing the objectives and activities.

Point 2: Getting to know the students better

AI can create exercises based on students' passions (for example, sport, music).

This motivates students and helps them to work for longer.

Point 3: Helping all students

The CUA is a way of thinking about lessons for all students.

AI can provide simple texts, audio, images, or games for learning.

Point 4: Differentiating learning

AI can adapt the content, activities, and required outputs.

Each student can learn with a different path towards the same goal.

Point 5: Establishing rules for AI in schools

The guide states that clear rules are needed for using AI in schools.

These rules should be discussed with students, teachers, and parents.