Generative AI is entering classrooms quickly, but teachers still lack clear, curriculum‑aligned guidance for using it responsibly. This report shows that existing lower secondary STEM curricula already offer strong, well‑guarded opportunities for GenAI‑enhanced inquiry and critical thinking, giving GenSTEAMers a solid foundation for practical, classroom‑ready work.
GenSTEAMers is an international Erasmus+ cooperation project that explores how Generative Artificial Intelligence (GenAI) can be meaningfully integrated into STEAM education to support both teaching and learning, with teachers’ experience, needs, and classroom realities as the starting point for innovation. Within this wider, teacher‑centred project, this report is the first major research output: a cross-country analysis of how lower secondary STEM curriculum goals can be responsibly connected to GenAI, with a particular focus on creativity and critical thinking in classroom practice. It is designed to give teachers, school leaders, and project partners a clear picture of where GenAI can support learning, and what safeguards are needed to keep student reasoning at the centre.
Before any classroom tools, scenarios, or training materials are created, the consortium needed to understand how existing lower secondary curriculum objectives in Ireland, Greece, Serbia, and Spain (Galicia) already support inquiry, modelling, data handling, and ethical discussion, and where GenAI could add value and foster creativity and critical thinking, in a carefully supervised way. This resulting report provides this evidence base, positioning curriculum‑anchored opportunities and risks so that later design and piloting phases have a clear, shared reference point.
The work was carried out between October 2025 and January 2026 by a partnership of practising STEM teachers, curriculum specialists, and education researchers in the four participating countries. Using a shared data collection template and codebook, the team reviewed and coded 153 curriculum entries and synthesised 136 lower secondary objectives (ISCED Level 2), looking at subject area, objective type (such as conceptual knowledge, procedural skills, design/engineering, inquiry, data handling, modelling, ethics), cross-cutting competences, possible GenAI uses, risk levels, teacher guardrails, and implementation status.
A stage-based inclusion rule ensured that only objectives explicitly labelled as lower secondary (12-15yrs) were included in the cross-country thematic analysis, with 17 other entries retained for transparency but excluded from the synthesis. Through this collaborative process, the partners identified six reusable pedagogical alignment patterns and nine cross-country themes and practices that together describe lesson structures and project framings teachers can recognise and adapt in their own classrooms.
Methodologically, the report focuses on envisaged rather than currently observed classroom use. Across the four systems, student-facing GenAI use was still very limited, so the mapped AI roles reflect professionally plausible, curriculum-aligned scenarios rather than established routines. No objective in the dataset was coded as Established Practice; implementation remains exploratory or emerging, which is why GenAI is treated as a co‑pilot to be questioned and verified, not as an autopilot for solving STEM tasks. Teachers consistently highlighted Accuracy/Validity and Over‑Reliance as dominant risks, shaping the emphasis on verification steps, critique protocols, and visual or data-based checking throughout the patterns and themes.
Within GenSTEAMers, the purpose of this report is threefold. It shows that there is already strong alignment potential between existing lower secondary STEM curricula and responsible GenAI integration, without major syllabus changes, directly informing the design of classroom‑ready teaching scenarios. It distils this potential into practical design language, giving the consortium a shared vocabulary for structuring GenAI‑enhanced activities. This includes concept clarification, inquiry design, pattern identification, and model exploration, which will feed into the GenSTEAMers Action Plan Kit and the digital Teacher Hub and it sets out the design principles and constraints for the forthcoming Action Plan Kit, specifying the kinds of GenAI roles, risk mitigations, and assessment focuses that classroom resources must embody if they are to foster both STEAM competences and critical AI literacy. In this way, the report is not only a research output but the structural foundation from which GenSTEAMers will build and iteratively refine concrete, classroom‑ready tools, project exemplars, and professional development materials, in close collaboration with teachers and schools.
Reports are available in Englsih, Serbian, Greek, and Spanish languages and may be found in the Results section of this website.