From 2026, the Italian school system enters a phase of reorganization that directly affects lower secondary: on the one hand, the issue ofLatin in middle school 2026(with greater flexibility and optionality), and on the other, an acceleration on digital skills withmandatory STEM computer sciencestarting already in primary school. For teachers, this means rethinking prerequisites, progressions, and assessment, while at the same time equipping themselves with concrete tools to design more modular and inclusive teaching. In this scenario, AI can become an ally forteacher training 2026and for building interdisciplinary pathways that are sustainable over time.
What changes from 2026: optional Latin in middle school and mandatory computer science from primary school
Thenew national curriculum guidelines(2026 horizon) aim for two combined effects: greater personalization of language-and-humanities pathways in lower secondary and a structural strengthening of the digital area. In practice, any study of Latin in middle school becomes more tied to the school’s choices and the pathway options, while computer science is treated as basic literacy, not as an occasional “project.”
The knock-on effect on upper secondary is immediate: incoming classes will have a more heterogeneous profile. Some students will already be familiar with basic morphology and vocabulary, others not; at the same time, many will have practiced computational thinking, conscious use of data, and first coding experiences. For departments, this means clarifying what theexpected competenciesare on entry and how to remediate or strengthen them without slowing down the pathway.
A key point: if computer science becomes more solid vertically, expectations also change around research, production of assignments, source management, and digital citizenship. It’s not just “knowing how to use tools”: it’s buildingteachers’ digital skillsand students’ skills to work with data, texts, arguments, and authentic problems.
who we are
Heterogeneity on entry requires an update of the vertical curriculum: it’s not enough to “add hours” or insert an initial test. It is necessary to define2026 is not asking teachers to “do more,” but to do things in a more integrated way: clear vertical curricula, competency-based assessment, and pathways that bring together languages, logic, and digital. With adequate tools and shared planning, Latin and STEM can become two complementary levers to improve students’ method, precision, and autonomy. If you want to start right away with a guided pilot, you can alsosign up for free
On the assessment side, it’s worth separating clearly:knowledge(e.g., paradigms, syntax, basic commands),skills(analysis, guided translation, writing a simple algorithm) andcompetencies(applying to new contexts, arguing choices, documenting a process). With transparent rubrics, the student understands what they are improving and the class council can align criteria across subjects.
Operationally, many schools are adopting three simple but effective moves:
- Short diagnostic entry tests (not selective) on Latin and digital skills, with formative feedback.
- “Optional” realignment modules (micro-units) to fill gaps without weighing down the whole class.
- Evidence-based assessment: products, processes, and metacognitive reflections, not only timed tasks.
Latin + STEM: interdisciplinary teaching models (UDA, authentic tasks, PBL) ready for 2026
The intersection between Latin and STEM is not an artifice: both train analysis, formalization, attention to structure, and error control. The key is anAI interdisciplinary teaching approachthat doesn’t “mix everything,” but builds UDAs with measurable objectives and assessable final products. Below are three replicable models for the first two years (adaptable to lyceums and technical institutes).
1) UDA “Morphology as a system”: from paradigm to rule
Objective: recognize regularities and exceptions in a system of forms. Activities: analysis of Latin word families, construction of transformation tables (root, stem, ending), comparison with rewriting rules in computer science. Final product: a “rule manual” (text + examples) and a simple script/pseudocode that, given a lemma, suggests possible forms (even if only partially). Assessment: rubric on accuracy, explanation of the process, quality of examples, and handling of exceptions.
2) Authentic task “Etymologies and data”: words, frequencies, contexts
Objective: use data to support a linguistic argument. Activities: collection of a small corpus (short texts provided by the teacher), manual or guided extraction of keywords, classification by semantic fields, comparison with derived Italian terms. Final product: an argumentative report with frequency tables and a commented etymological glossary. Assessment: coherence of the hypothesis, traceability of sources, quality of categorization, and clarity of exposition.
3) PBL “Translating is problem-solving”: strategies and debugging
Objective: develop checking and revision strategies. Activities: step-by-step translation of a passage with “checkpoints” (sentence analysis, lexical choices, syntactic rendering), introducing the idea of debugging: identify the error, hypothesize the cause, test a correction. Final product: commented translation with a decision log (why I chose this construction; which alternative I discarded). Assessment: effectiveness of strategies, ability to justify choices, improvement between first and second version.
These pathways work well if coordinated within the class council: Italian for argumentation, mathematics for formalization, technology/computer science for the operational part, history for context. The advantage is twofold: fragmentation is reduced and continuity between humanities and scientific competencies is made visible.
How StudierAI supports teachers: planning, differentiation, and assessment with AI tools
To tackle the 2026 changes, you need planning time, coherence among teachers, and differentiated materials. This is where an environment likeStudierAIcan become practical support: it doesn’t replace the teacher, but speeds up repetitive phases and makes it easier to keep objectives, activities, and assessment aligned.
In particular, AI is useful in four key moments:
- UDA planning: generate a structured draft (milestones, objectives, prerequisites, phases, timing) and then adapt it to the real context of the class.
- Differentiation: create variants of the same task (basic/intermediate/advanced), with scaffolding and more guided instructions for those who need to realign.
- Assessment: build rubrics consistent with competencies and products (process, accuracy, autonomy, argumentation), as well as tests with explicit criteria.
- Monitoring: collect evidence and descriptors to observe progress on transversal competencies (study method, collaboration, digital citizenship).
An often underestimated aspect ofteacher training 2026is sustainability: tools and procedures must be replicable, shareable, and updatable. If you want to try a faster workflow for UDAs, tests, and rubrics, you canstart for freeand build the first materials starting from your disciplinary and interdisciplinary objectives.
Finally, when AI is introduced at school, clarity on responsibility and conscious use is essential: privacy, transparency of assignments, education in the critical use of sources. To learn more about the project’s approach and principles, you can consultwho we are.
2026 is not asking teachers to “do more,” but to do things in a more integrated way: clear vertical curricula, competency-based assessment, and pathways that bring together languages, logic, and digital. With adequate tools and shared planning, Latin and STEM can become two complementary levers to improve students’ method, precision, and autonomy. If you want to start right away with a guided pilot, you can alsosign up for freeand set up your first interdisciplinary UDA in just a few iterations.
