Thenew national guidelines 2026push schools toward a more flexible curriculum, centered on skills, inclusion, and responsibility in the use of technology. For teachers, the question is not “whether” to use AI, buthow to use it in a didactically sound way: to design, differentiate, document, and assess without losing control of the educational process.
In this article you’ll find an operational framework: what changes from 1 September 2026, a replicable 6-step workflow to build apersonalized curriculum with AI, ready-to-use tools for upper secondary school, and concrete guidance on assessment, inclusion, and data protection. The goal is to turnpersonalization learning artificial intelligenceinto sustainable practices, consistent with pedagogical evidence and professional ethics.
What the New National Guidelines 2026 change: personalization, skills, and ethical AI
From 1 September 2026, the approach of thenew national guidelines 2026more strongly emphasizes three pillars:personalization, development ofskillsandresponsible use of AIas a cultural environment and a tool. In practice, planning is no longer exhausted by “whole-class” programming, but requires differentiated pathways, transparent criteria, and documentation of instructional choices.
On the pedagogical level, the direction aligns with well-established evidence: personalization works when it combinesclear objectives, frequent feedback, targeted remediation practices, and authentic tasks. AI can strengthen these elements, but it does not replace them: it speeds up analysis, differentiation, and material production, while leaving instructional leadership to the teacher.
Operational implications for departments and class councils:
- Planning: move from “content to cover” to maps of skills, prerequisites, and mastery levels, with alternative activities and flexible timing.
- Assessment: shared rubrics and criteria, attention to processes (strategies, revisions, sources) as well as final products.
- Ethical AI: transparency with students and families, rules of use, data protection, and activities that develop critical thinking and information literacy.
In short: AI is useful if it is embedded in intentional planning. Inai in upper secondary school teaching, this means structuring tasks that require personal reworking, reasoning about sources, and coherence checks, instead of automatically generated “perfect answers.”
From planning to a personalized curriculum: a 6-step workflow with AI
Apersonalized curriculum with AIdoes not come from a prompt like “write me a unit.” It works better as a workflow: essential instructional inputs, generation of alternatives, teacher selection, and then monitoring. Below is a 6-step process, replicable for learning units and individualized pathways.
1) Define the expected outcomes (skills and criteria). Start with 2–4 observable outcomes: what should the student be able to do? Already link 1–2 quality indicators. Here AI can help translate general goals into operational formulations and propose mastery levels.
2) Map prerequisites and typical obstacles. Gather quick evidence: frequent mistakes in tests, classroom observations, results of entry assessments. AI is useful for turning these data into an “obstacle map” and suggesting targeted micro-interventions (alternative explanations, graded exercises, examples).
3) Design tiered activities (core + extensions + remediation). Structure a sequence with: essential activities for everyone, enrichment tasks for those who move faster, and remediation pathways for those with gaps. AI can generate variants of the same task with different degrees of support (scaffolding, guided examples, reduced cognitive load).
4) Define evidence and products. To avoid “invisible” personalization, decide what evidence you will collect: mini-quizzes, written work, oral explanations, concept maps, revision logs. AI can propose observation grids and quality checklists, but the choice must remain consistent with the subject and the class context.
5) Build a lightweight (non-bureaucratic) individual plan. For some students, a realpersonalized study plan studentsis needed: priority objectives, recommended activities, timing, formative checks, and criteria. AI helps turn the diagnosis (strengths/weaknesses) into a sustainable study sequence, with weekly micro-goals and coherent materials.
6) Monitor and adjust (short cycle). Personalization is effective if iterative: collect evidence, give feedback, update activities. AI can summarize progress and suggest “next steps,” but the teacher always validates: what has truly improved? what is only appearance? which strategies worked?
Practical tip: to make the workflow sustainable, prepare an input template (objectives, prerequisites, time constraints, class level, specific needs) and reuse it. This is whereai tools for teachers 2026come into play: not just content generators, but assistants for planning, differentiation, and documentation.
“Ready-to-use” AI tools for teachers: summaries, flashcards, quizzes, oral simulations, planner
To integrate AI without distorting teaching, it helps to think in phases: before the lesson (preparation), during (interaction and feedback), after (study and consolidation). Here are 5 tool families with examples of prompts and expected outputs, particularly useful inai in upper secondary school teaching.
1) Summaries and controlled simplification. Before the lesson: create a 10-minute “pre-reading” with graded vocabulary. During: check understanding with targeted questions. After: provide a summary in two versions (standard and simplified) for independent study. Expected output: short texts with keywords, examples, and a minimal glossary.
Example prompt for AI (to be validated): “Summarize this paragraph in 120 words for a third-year class, highlight 5 keywords and explain 3 difficult terms with a concrete example.”
2) Flashcards and spaced review. After the lesson: generate flashcards in question/answer pairs, also including distractors and reminders of common errors. Expected output: a set of 15–25 cards per unit, with progressive difficulty and an indication of what to memorize vs what to understand. This supports autonomy and reduces teacher prep time.
3) Formative quizzes and immediate feedback. Before: a short entry test to map prerequisites. During: a mid-lesson “stop&check.” After: a consolidation quiz with explanations of errors. Expected output: multiple-choice questions, true/false with justification, short-answer items, with key and rationale. Good practice: have AI include the justification for the correct answer and why the alternatives are plausible but wrong.
4) Oral simulations and Socratic dialogues. During and after: train presentation, argumentation, and metacognition. Expected output: an oral-exam outline with graded questions (recall, application, connection, evaluation), and a follow-up “script” that asks for examples, counterexamples, definitions, and interdisciplinary links. This is particularly useful for students who struggle to organize their speech.
5) Planner and study plans. After: turn objectives and constraints into a realistic calendar. Expected output: daily/weekly micro-goals, estimated times, priorities, and self-check moments. This is wherepersonalization learning artificial intelligencebecomes concrete: not “study more,” but “review these 8 concepts with 12 flashcards, then do 10 quizzes, then record a 2-minute explanation and compare it with the rubric.”
Methodological note: to avoid generic outputs, always provide context (class, level, time, objectives), constraints (number of items, duration, format), and quality criteria (e.g., “include 3 typical errors and explain how to recognize them”). AI makes production fast, but quality depends on teacher direction.
Assessment, inclusion, and data protection: rubrics, transparency, and prevention of misuse
Personalization requires coherent assessment: if I differentiate activities and supports, I must make clear what remains common to everyone (standards) and what changes (pathways). The most robust solution is arubricwith stable criteria and descriptive levels. AI can help write clear descriptors, but the rubric must be discussed within the department and “taught” to students (assessment as learning).
Transparency and proper use of AI: define a 5-line class policy, with examples. For instance: when AI is allowed (brainstorming, revision, exercises), when it is forbidden (in-class tests), and what must be disclosed (prompts used, reworked parts, sources). This reduces conflict and turns prevention into education for digital citizenship.
Inclusion: AI can increase accessibility (simplifications, audio, examples, controlled translations), but it must be used carefully so as not to create dependency or lower expectations. Good practices:
- Offer multiple representations of the same content (text, examples, guided exercises) while keeping essential objectives unchanged.
- Always require personal reworking: explanation in one’s own words, an original example, a link to a real case.
- Use metacognitive checklists: “What didn’t I understand? What strategy did I use? What evidence do I have that I can do it now?”
Data protection and traceability: avoid entering unnecessary personal data into external tools (names, diagnoses, sensitive information). Prefer anonymous identifiers or generic profiles. Require students to cite sources and save a brief “process note” (main prompts, revisions, choices). This also helps prevent misuse: you don’t need to “hunt” AI—you need to make the process assessable.
How StudierAI can support personalized planning (and save teachers time)
In a context where personalization becomes structural, teachers’ time is the scarcest resource.StudierAIcan support planning and study guidance while keeping the focus on objectives, evidence, and criteria. The idea is not to “automate teaching,” but to speed up what is repetitive (materials, variants, exercises) and strengthen what matters (coherence, feedback, monitoring).
Here’s how it can help in concrete ways, in line with thenew national guidelines 2026and with a professional approach to personalization:
- Create personalized study plans: starting from objectives, time, and difficulties, it supports building a pathway with micro-goals and priorities, useful as a real work plan for the student.
- Generate consolidation materials: summaries, flashcards, and quizzes consistent with the content and level, reducing prep time and increasing the frequency of formative feedback.
- Simulate oral exams: graded questions and follow-ups to train presentation and argumentation, with attention to typical errors and threshold concepts.
- Monitor progress: help the student see what has been done, what remains, and where to focus, fostering self-regulation and continuity in study.
If you want to see whether this approach can truly lighten your planning, you canstart for freeand test the creation of materials and pathways on one of your units. To learn more about the educational vision and the project context, you’ll find more information on theabout uspage.
Operational closing: whatever tool you choose, keep three quality anchors:explicit objectives,observable evidenceandtransparent criteria. Under these conditions, AI becomes a reliable accelerator: it helps build a more equitable, sustainable curriculum that is truly student-centered.
