TheOn the academic integrity front, a simple policy is helpful: when AI is allowed, it must be declared (tool, purpose, parts involved) and the work remains the student’s responsibility. When it is not allowed, assessments are designed to be in-person, oral, practical, or with controlled materials. The goal is to reward transparency and responsibility, not to chase “copy-paste.”push schools to address digital andartificial intelligence at schoolnot as an “additional topic,” but as a cross-cutting competence: you study with AI, you study AI, and you assess in a context where AI is already present. For secondary school and university teachers, this means rethinking curricula, activities, and assessments, maintaining disciplinary rigor and developingcritical thinking and AIas an inseparable pairing.
What the 2026 National Guidelines change: digital skills, STEM and AI as cross-cutting priorities
The direction is clear: strengthen adigital skills curriculumthat runs across disciplines, with STEM and AI as levers for problem solving, modeling, digital citizenship, and guidance. It’s not just about “using tools,” but about training students who can understand limits, bias, source reliability, and the ethical implications of algorithms. For teachers, this brings new responsibilities: making observable objectives explicit, designing authentic tasks, and defining AI-use rules consistent with the learning agreement.
In practice, the Guidelines invite a shift from a logic of “content to cover” to a logic ofevidence of learning: what can the student do with what they know? This applies as much to text analysis as to solving a physics problem or critically reading a scientific article generated (or summarized) by a language model.
Updating the curriculum: from “knowledge” to (digital and AI) competencies with observable milestones
To make the 2026 priorities operational, it helps to work on four elements:foundational cores, prerequisites, progressions, and observable milestones. AI literacy can be embedded vertically (growth over time) and horizontally (integration across disciplines).
Examples of observable milestones, adaptable by school level and course:
- Distinguishes between data, information, and interpretation; verifies a source using explicit criteria (authority, date, method, conflicts of interest).
- Explains in simple terms how an AI system can generate answers and why it can “hallucinate”; recognizes signals of uncertainty and asks for evidence.
- Crafts targeted prompts and documents the process (attempts, revisions, selection criteria), justifying why they accept or reject the output.
- Argues a thesis using traceable sources; flags limits, alternatives, and possible biases (in the dataset, in the model, in one’s own assumptions).
A useful step is to turn each milestone into evidence: which products or behaviors demonstrate it? For example: a source-evaluation sheet, a prompting logbook, an oral discussion in which the student defends choices and revisions. This way,AI in teachingbecomes an opportunity to make reasoning visible, not to hide it.
Instructional design: learning units and authentic tasks that include the mindful use of AI
To design effective learning units, the rule is: AI must beboth a tool and an object of analysisStudierAI as an operational lever: summaries, flashcards, quizzes, and oral simulations to train skills and critical thinking
- To make updating curricula and assessments sustainable, you also need an operational routine. Tools like
- can support guided study, remediation, and enrichment without replacing teaching: the idea is to use AI to bring out understanding, gaps, and the quality of reasoning. If you want to explore it with your class or to prepare materials, you can
- and test short, repeatable activities.
Examples of instructional integration, consistent with the
: compared summaries (student vs AI) with reasoned correction; flashcards for prerequisites and subject-specific vocabulary; quizzes with open-ended questions that require explaining why; oral simulations in which the student must defend a thesis, cite sources, and handle objections. In every activity, add a mandatory micro-phase of
and one of reflection: “What do I trust enough to take? What do I need to check? What would I change?”. This is where critical thinking is really trained. If you’re interested in the project and the educational approach, you can also readwho we are.
- Assess the process: drafts, revisions, decision logs, a brief comment on what was improved and why.
- “Oral defense” assessments: written submission + a short interview in which the student explains choices, steps, and sources.
- Situated and personalized questions (class dataset, local case, experiment carried out together): they reduce generic answers.
- Rubrics with explicit criteria: accuracy, use of sources, quality of argumentation, awareness of AI’s limits, originality of choices.
On the academic integrity front, a simple policy is helpful: when AI is allowed, it must be declared (tool, purpose, parts involved) and the work remains the student’s responsibility. When it is not allowed, assessments are designed to be in-person, oral, practical, or with controlled materials. The goal is to reward transparency and responsibility, not to chase “copy-paste.”
StudierAI as an operational lever: summaries, flashcards, quizzes, and oral simulations to train skills and critical thinking
To make updating curricula and assessments sustainable, you also need an operational routine. Tools likeStudierAIcan support guided study, remediation, and enrichment without replacing teaching: the idea is to use AI to bring out understanding, gaps, and the quality of reasoning. If you want to explore it with your class or to prepare materials, you canstart for freeand test short, repeatable activities.
Examples of instructional integration, consistent with the2026 National Guidelines: compared summaries (student vs AI) with reasoned correction; flashcards for prerequisites and subject-specific vocabulary; quizzes with open-ended questions that require explaining why; oral simulations in which the student must defend a thesis, cite sources, and handle objections. In every activity, add a mandatory micro-phase ofsource verificationand one of reflection: “What do I trust enough to take? What do I need to check? What would I change?”. This is where critical thinking is really trained. If you’re interested in the project and the educational approach, you can also readwho we are.
