Thenew National Guidelines 2026are not just a “paper” update: they affect planning, assessment, materials, and the evidence a teacher must be able to collect. In parallel, the adoption ofAI tools for teachersis becoming a factor of efficiency and quality, provided professional control, transparency, and instructional coherence are maintained. In this article you’ll find an operational method to turn objectives and learning outcomes into units, ready-to-use lessons, and assessments, using AI as a planning assistant (not as an “autopilot”).
The guiding idea: start from what changes (objectives, competencies, inclusion, and assessment) and build a workflow in which AI produces drafts, alternatives, and variants, while the teacher sets criteria, selects, adapts to the context, and documents. It’s an approach consistent with the literature on backward design, constructive alignment, and assessment for learning: clear objectives → targeted activities → coherent assessment evidence.
What the New National Guidelines 2026 change and what it means for planning
While awaiting the final framework, the operational horizon from1 September 2026implies for schools a realignment among: competency targets, core disciplinary concepts, cross-curricular dimensions (key competencies, citizenship, digital), inclusion, and assessment. For2026 instructional planningthis translates less into “adding content” and more into making the decision chain explicit: why do I choose these activities? What evidence do I collect? How do I ensure accessibility and progression?
To work in a robust way, it helps to break down every expected change into four operational choices, independent of the subject:
- Units and progression: define core concepts, prerequisites, micro-objectives, and a progression of difficulty (from guided to independent).
- Competencies and performances: describe what the student “can do” in observable tasks (products, procedures, arguments, transfer).
- Assessment: criteria and rubrics consistent with the objectives, with clear indicators and descriptive levels; balance between formative and summative.
- Materials and accessibility: multi-channel resources, linguistic simplifications, compensatory tools and strategies for SLD/SEN, with attention to cognitive load.
This framework is consistent withconstructive alignment: objectives → activities → assessment. It is also the basis for using AI sensibly: if criteria are explicit, AI can help generate variants of activities and tests, but it cannot decide in place of the teacher what is “good” for that class and that context.
2026 instructional planning with AI: from objective analysis to ready-to-use units and lessons
An effective AI workflow doesn’t start with the request “make me a lesson,” but with an essential input sheet. The more structured the input, the more controllable the output. Below you’ll find a replicable model for UDAs and lessons, useful with a 2026 perspective.
1) Define instructional constraints (5 minutes). Write: class, subject, duration, prerequisites, group profile (heterogeneity, SLD/SEN), objectives in observable form, and expected evidence. This is the core of2026 instructional planning: objectives are not “topics,” but performances.
2) Ask AI for a map of units and lessons (10 minutes). Useful prompt: “Propose a UDA of 4–6 lessons with objectives, activities, timing, materials, inclusive strategies, and a final authentic task. Keep coherence with competency targets and transparent assessment.” AI produces a draft: you check progression, feasibility, and adherence to your criteria.
3) Apply backward design (10 minutes). Start from the final task: what product or performance demonstrates the competency? Then work backward: what practice, models, feedback, and recovery moments are needed? This reduces the “nice but disconnected lessons” effect and supports formative assessment.
4) Generate rubrics and criteria, then simplify (10 minutes). Ask AI for a 4-level rubric with descriptive indicators. Immediately after, ask for a “student-friendly” version (accessible language) and a “teacher” version (more analytical). The key step is revision: remove redundant indicators, check that levels are progressive, and that wording is unambiguous.
5) Inclusion and accessibility: create variants (10 minutes). AI is very useful for producing: simplified instructions, glossaries, text-based concept maps, easy-to-read versions, graded exercises, and compensatory tools. Here the instructional rule is:same objective, different pathway. Avoid instead “lower objectives” unless justified by an IEP/individual plan: better to work on supports, time, channels, and feedback.
6) Document decisions (5 minutes). To be consistent with a 2026 framework oriented to competencies and transparency, keep: prompts used, versions of materials, criteria, and rationales for choices. It’s not bureaucracy: it’s professional traceability and protection.
Assessments with AI: quizzes, authentic tasks, and coherent rubrics (without losing control and transparency)
Theassessments with AIwork when AI is used to increase quality and variety, not to “mass-produce” misaligned tests. The point isn’t having more questions, but better questions: valid with respect to objectives, calibrated for difficulty, and with clear marking criteria.
A practical method in 4 steps:
- Define the objectives × levels matrix: for each objective indicate at least two performance levels (basic/advanced) and the type of evidence (selection, production, argumentation, application).
- Generate varied items: ask AI for a balanced set (e.g., 30% recall, 40% understanding/application, 30% transfer) including plausible distractors and feedback for the typical error.
- Anti-bias and accessibility review: check language, cultural examples, implicit references, length of instructions; ask AI for an “easy-to-read” version and one with disciplinary vocabulary, keeping the assessed construct unchanged.
- Traceability: keep the marking grid, answer key, rationales for choices, and final version; if you modify an item, note why (ambiguity, excessive difficulty, content not covered).
For authentic tasks, AI is useful in generating realistic scenarios and constraints (data, sources, roles), but validity depends on how you build the criteria. A good check is this: every rubric criterion must be linked to an observable action (e.g., “selects relevant sources and justifies them” is observable; “has understood” is not).
Finally, transparency toward students and families: state what you assess, with which criteria, and how AI will be used (if you use it) in preparing materials. Transparency reduces conflict, increases the perception of fairness, and supports motivation.
Flashcards for teachers and study materials: review, remediation, and enrichment with AI
Theflashcards for teachersare not “little games”: if well designed, they support remediation and consolidation through active recall and spaced repetition. AI speeds up the production of coherent sets, but quality depends on how you define the conceptual scope and the typical errors you want to catch.
Recommended workflow to create study materials with AI:
- Define core concepts and boundaries: 10–15 key concepts, 10 examples, 10 non-examples. Ask AI to propose concept→definition and definition→concept flashcards, including frequent mistakes and well-reasoned “traps.”
- Gradation: ask for three-level sets (basic, intermediate, advanced) with constraints on answer length and vocabulary complexity.
- Reducing cognitive load: ask AI to break long explanations into micro-steps, add signals (linking words, lists), and propose worked-out examples before independent exercises.
- SLD/SEN: generate alternatives with high-legibility fonts (if предусмотрено in the materials), shorter sentences, an integrated glossary, and step-by-step instructions. For the same competency, you can offer multiple channels: guided oral, outlines, hands-on manipulation, concrete examples.
A particularly effective use is “targeted review”: after an assessment, you select 3–4 typical errors that emerged and ask AI to create mini-sets of flashcards and exercises focused on those errors, with immediate feedback. This way assessment becomes truly informative and guides subsequent teaching.
AI instructional planner and tools: how StudierAI can support lessons, assessments, and materials aligned with 2026
When the goal is to align planning, materials, and assessment with thenew National Guidelines 2026, anAI instructional plannercan make a difference especially on three fronts: coherence (the same objectives throughout the whole pathway), speed (drafts and variants), and documentation (traceability of choices). Tools likeStudierAIcan support teachers’ work if used with a “co-planning” logic: you define criteria and context, the tool generates and organizes reusable outputs.
Here are some concrete use cases, easily transferable to everyday practice:
- UDA and lesson planning: generating lesson sequences with observable objectives, timing, materials, and feedback strategies; creating variants for classes with different levels.
- Coherent assessment: drafts of rubrics, grids, level descriptors, and “student-friendly” criteria; generating performance examples to clarify what is meant by basic/intermediate/advanced.
- Tests and remediation: creating items for quizzes, structured tests, and authentic tasks; producing targeted remediation sets on the typical errors that emerged (with explanations and graded exercises).
- Study materials: flashcards, leveled summaries, glossaries, prompts for oral exams, text-based maps, and self-assessment sheets to make students more autonomous.
Best practices for integrating the tool into teachers’ work: (a) always work with a stable “criteria sheet” for subject/class; (b) have it generate at least two alternatives and choose consciously; (c) keep prompts and versions as documentation; (d) always verify content, examples, and data. If you want to experiment lightly, you canstart for freeand build your first planning and assessment workflow; to learn about the project and the educational approach behind it, you’ll find more information on theabout uspage.
Looking ahead to 2026, the distinctive competency won’t be “using AI,” but knowing how todesign and assess with explicit criteria, leveraging AI to reduce dead time and increase the quality of variants (inclusion, levels, feedback). If you want to put together a first set of reusable UDAs, assessments, and materials, you can alsosign up for freeand start from a real unit of yours: it’s the fastest way to turn principles and keywords into sustainable practices.
