Luxembourg - AI in Schools Guidelines

Overview

KI Kompass is a consultation-stage strategic proposal (version of 6 October 2025) by the Ministry of Education, Childhood and Youth (MEN) to orient, support and regulate responsible AI use in Luxembourg schools. The proposal defines an education-centred approach that privileges human competencies and pedagogical aims before technological adoption. Central components include an online platform (ki-kompass.lu) providing validated AI tools for classroom use, a broad continuous professional development catalogue developed by the national training institute, communities of practice for peer exchange, and a helpdesk for pedagogical and legal queries. The document is grounded on national survey data (4,037 pupils, ~240 teachers) and proposes a staged, maturity-based model for progressive AI integration in learning and assessment. The MEN has set a consultation timeline ending December 2025 and plans to finalise operational guidelines in January 2026; the proposal therefore remains under review pending stakeholder feedback and scientific monitoring. The strategic aim is dual: (1) to equip students with AI literacy and critical competence and (2) to ensure AI is used only as a pedagogical co‑agent, supported by validated technical, data protection and governance safeguards. Many elements of KI Kompass directly reference EU-level regulatory obligations and Luxembourg supervisory bodies, and the platform is intended as a national coordination instrument to reduce fragmentation in tool use and to advance consistent, safe practices across the school system.

Definitions

The KI Kompass proposal supplies working definitions tailored to schools: "Artificial Intelligence (AI)" is defined as systems that analyse data, learn and provide results such as classifications, predictions, recommendations or generated content. "Generative AI" denotes systems that produce novel content based on prompts. "Augmented/extended intelligence" or "co-agent" describes human–AI collaboration that supports learning without replacing student cognition. The framework also defines an "AI-competent user" (students or staff with graded competencies), "validated AI tool" (an application assessed and approved for school use through a ministerial validation process), and "maturity levels" (a staged model governing permissible autonomy in tool use tied to age and digital maturity). The document clarifies terms supporting integrity policies (e.g., "use with attribution", "unauthorised delegation", "AI-assisted work") and technical concepts relevant to procurement and privacy (e.g., data minimisation, pseudonymisation, model provider obligations).

Governance and Institutional Framework

The governance model proposed combines central coordination by the Ministry of Education with operational implementation by SCRIPT (the Service for coordination of pedagogical and technological research and innovation). The document designates MEN/SCRIPT as the primary sponsors of the KI Kompass platform and the national contact point for schools; it envisages scientific accompaniment by external partners and systematic consultation with school directors, teacher unions and representative bodies. Relevant national oversight authorities (including the Commission nationale pour la protection des données, CNPD) are identified as stakeholders for data protection and regulatory compliance. Institutional responsibilities are allocated as follows: MEN issues strategy and validation criteria; SCRIPT manages platform content, training and communities of practice; school directors implement local policies consistent with the maturity model; teachers exercise classroom-level pedagogical judgement; and national regulators (CNPD and other designated competent authorities under the EU AI framework) retain enforcement and supervisory powers. The proposal also recommends an internal documentation regime for schools registering AI uses and a central registry of validated tools maintained on the platform. See the MEN press release and downloadable strategy documents for organisational details (MEN news release, ki-kompass.lu).

Key Focus Areas

The KI Kompass strategy concentrates attention on several interlinked domains: (1) Pedagogy and student competencies — a maturity-based progression where primary education focuses on human skills and minimal direct AI use, lower secondary covers "learning about AI" and risks, and upper secondary supports responsible, supervised "learning with AI"; (2) Teacher capacity-building — a nationwide catalogue of more than 80 training modules, thematic learning pathways and school-level bespoke trainings; (3) Tool validation and provisioning — a curated set of validated, secure AI tools provided free via the national platform to reduce unregulated third-party adoption; (4) Academic integrity and assessment — clear guidance on evaluation practices, alternatives to punitive detection methods, and internal sanction frameworks for misuse; (5) Data protection, privacy and cybersecurity — compliance with the EU AI framework and CNPD guidance emphasizing data minimisation, documentation and informed consent where applicable; (6) Monitoring, research and pilots — scientific monitoring, planned surveys (KI-Umfrage 2026), national pilot projects and an evaluative cycle to iteratively refine policy; and (7) Public consultation and stakeholder dialogue — a mandated consultation process (ending Dec 2025) followed by a January 2026 operational conference for teachers and directors. The policy balances opportunity (personalisation, inclusion, time savings) with risks (cognitive outsourcing, bias, misinformation, student data exposure), and commits to human-centred safeguards in design and deployment.

Implementation Framework

Implementation is staged. Phase 1 focuses on teacher upskilling, platform launch and piloting of validated tools in selected schools. Phase 2 extends student-focused competence development and wider tool availability. The operational deliverables include: (a) the KI Kompass one-stop digital platform; (b) a ministerial validation procedure and registry for school-use AI tools; (c) a documented maturity model embedded in curricula (notably Digital Sciences); (d) school-level policies for procurement, privacy and incident reporting; (e) Communities of Practice to foster peer learning; and (f) a helpdesk for pedagogical and legal questions. The proposal mandates documentation of deployments (who uses the tool, purpose, data flows, safeguards) and suggests templates and checklists for schools to follow. The framework recommends periodic internal audits, an evidence-based evaluation plan with predefined KPIs (learning outcomes, teacher satisfaction, integrity incidents, privacy incidents) and alignment of procurement with national cybersecurity and data-protection standards. The MEN proposes public dissemination of guidance materials and a national conference for dissemination in January 2026 (official announcement).

Monitoring and Evaluation

Monitoring is both formative and summative. The strategy defines continuous monitoring via platform analytics, annual surveys (KI-Umfrage), pilot evaluations, and scientific accompaniment by designated research partners (OEJQS or equivalent). Evaluation metrics include pedagogical efficacy, digital maturity progression, incidence of integrity violations, data protection incidents, and equity impacts. The MEN plans a revised policy after stakeholder consultation and a follow-up national survey in 2026. Schools are expected to maintain logs of AI use, training records, and incident reports to feed national monitoring. The proposal emphasises transparency in reporting outcomes to inform iterative policy adjustments, to identify tools requiring delisting, and to recommend targeted interventions where risks exceed acceptable thresholds.

Penalties, Liability, and Appeals

The consultation draft outlines internal school-level disciplinary frameworks for breaches of academic integrity and misuse of AI tools and signals that more serious incidents (data protection breaches, unlawful high-risk AI uses) will be referred to national competent authorities. The document notes the interplay with EU-level obligations under the AI legal framework and Luxembourg data protection law; where statutory breaches occur, enforcement follows national law and supervisory authority procedures (including possible CNPD actions). The CNPD and the forthcoming national competent authority designations under the AI Act are the appropriate referral points for statutory violations. The KI Kompass proposal also commits to transparent appeal mechanisms for schools and individuals affected by administrative decisions and to documenting liability allocation for providers of validated tools via procurement contracts and service-level agreements.

Relationship to Other Instruments

KI Kompass explicitly situates itself alongside existing instruments: Luxembourg’s education curricula (Digital Sciences), SCRIPT operational responsibilities, national data protection rules administered by the CNPD, and the EU AI regulatory framework. The strategy is designed to complement — not replace — statutory rules. It provides operational guidance to bring school practice in line with legal duties (e.g., documentation, AI literacy obligations, data‑protection impact assessments where required) and to inform national transposition of EU-level processes for designation of competent authorities and market surveillance. The proposal references existing teacher-training initiatives and Erasmus projects (AI4T and similar) as inputs and foresees integrating outputs from these initiatives into platform materials.

International Alignment

KI Kompass emphasises alignment with the EU AI regulatory approach and related guidance, including AI literacy recommendations from the European Commission and CNPD thematic materials. The strategy notes anticipated interactions with EU-level market surveillance, standardisation work and the designation of national competent authorities under the AI Act. It positions Luxembourg’s platform and validation procedures as potentially compatible with cross-border initiatives and pilot projects, and it invites (inter)national collaboration to exchange evidence, validation approaches and good practices. The MEN explicitly references EU instruments and CNPD guidance in the consultation materials to ensure interoperability and legal coherence.

Implementation Timeline

Milestone	Date
Official proposal published	2025-10-06
Consultation with school partners (end)	2025-12-31
Operational guidelines finalised (target)	2026-01-31
National teacher conference / dissemination	2026-01 (target)
Platform phased tool roll-out	2026 Q1–Q3
Follow-up national survey (KI-Umfrage)	2026 (planned)

Compliance Checklist

Checklist Item	Action / Evidence
Adopted school AI policy	Local policy document aligned with KI Kompass maturity model
Validated tool registry use	Screenshot / link to tool entry on ki-kompass.lu
Teacher training records	Training certificates / institute of education logs
Data protection assessment	CPIA / data minimisation record or CNPD advice
Documentation of classroom use	Log entries (purpose, users, data flows)
Integrity & assessment rules	Published school guidance & sanction protocol

Sources and References

Source	Type
KI Kompass: proposition of a strategic framework (MEN news release, 06/10/2025)	Primary Source
Proposition d’un cadre stratégique de l’intelligence artificielle à l’école (MEN, PDF, 06/10/2025)	Primary Source
Presentation: Orientation, practice and exchange on AI in school (MEN presentation PDF)	Primary Source
KI Kompass platform (official portal)	Primary Source
SCRIPT — Service for coordination of pedagogical and technological research and innovation	Primary Source
CNPD guidance on AI literacy and RIA obligations	Primary Source