Belgium - National AI Coalition

Overview

AI4Belgium is a national coalition and strategic roadmap that organises public, private and academic AI stakeholders in Belgium to champion responsible, human-centred AI deployment. Launched as a collaborative platform under the coordination of the federal Policy and Support service and with active participation from regional hubs and industry bodies, AI4Belgium serves three complementary roles: convenor and network, national observatory and monitoring platform, and practical guidance hub for trustworthy AI tools and processes. The initiative has produced public-facing materials, assessment tools and events and is referenced in EU-level reporting on national AI ecosystems (see EU Digital Decade country report for Belgium). AI4Belgium maintains online resources, including an ALTAI-based trustworthiness assessment tool and a coalition website used for outreach and publication of the roadmap and updates.

Definitions

For the purposes of AI4Belgium and the associated roadmap the following working definitions are used: "AI ecosystem" — the combined set of actors (public administrations, private companies, civil society, research centres) engaged in AI development, deployment and governance; "Trustworthy AI" — AI designed and implemented in accordance with ethical principles including human oversight, fairness, transparency and robustness; "Observatory" — a monitoring mechanism producing indicators, case studies and periodic updates on AI uptake and impacts; "AI sandbox / pilot" — controlled, time-limited environment where stakeholders can experiment with AI innovations under supervised conditions. These definitions are operational and intended to align with EU policy language while reflecting Belgian administrative practice.

Governance and Institutional Framework

AI4Belgium is governed as a coalition model rather than through statutory command-and-control. Coordination is led at federal level (FPS Policy & Support / BOSA) working with the three regional AI hubs (The Beacon in Flanders, Digityser in Brussels and Réseau IA in Wallonia), industry federations (notably Agoria), research organisations (e.g., imec) and civic actors. Institutional roles include: a national coordination secretariat to run the observatory and events; working groups for sectoral pilots (health, mobility, industry, public services); and a technical team maintaining open-source tools such as the ALTAI-derived assessment platform (hosted examples include ALTAI assessment tool). Governance relies on memoranda of cooperation, funding agreements and voluntary commitments — not new statutory authorities — though public procurement and grant-making are used to incentivise adoption of responsible AI practices. The model emphasises horizontal cooperation across jurisdictional levels and rapid dissemination of technical guidance to public administrations.

Key Focus Areas

AI4Belgium’s roadmap concentrates activity across several strategic pillars: capacity and skills (reskilling and upskilling programmes for public servants and industry, university and vocational training linkages); trustworthy AI adoption (promoting ALTAI-based assessments and transparency reports for public-facing systems); data and infrastructure (improving FAIR data availability, secure data sharing arrangements and access to HPC/cloud resources for research and SMEs); sectoral pilots and sandboxes (healthcare, transport, manufacturing and public administration pilots to test governance and safety mechanisms); SME and innovation support (targeted coaching, funding instruments and linking SMEs to Digital Innovation Hubs); ethics, rights and safety (embedding privacy-by-design, non-discrimination and human oversight in procurement and deployment); and international cooperation (alignment with EU regulatory pathways and participation in EU projects such as EDIHs and AI Factories). The coalition places special emphasis on lowering barriers for SMEs and public bodies to responsibly adopt AI by offering assessment tooling, best-practice templates and facilitated connections to research partners.

Implementation Framework

Implementation is organised through a mix of voluntary commitments, public funding calls, working groups, observatory reporting and technical tool deployment. The federal secretariat publishes guidance and operational recommendations for public procurement of AI and encourages public administrations to adopt ALTAI-style self-assessments and transparency measures. Regional partners run localized adoption programmes and training; industry partners provide case studies and sandbox partners host pilots. Funding channels include regional innovation grants, participation in EU funding streams and targeted federal calls aimed at bridging experimentation and deployment (particularly for SMEs). Technical implementation also includes the provision of open-source assessment tools (see the ALTAI tool hosted by the coalition) and shared datasets / sandboxes under strict privacy and security controls. The implementation model intentionally avoids duplicative regulatory requirements and instead prepares stakeholders for compliance with binding EU instruments while filling gaps in practical guidance and capacity.

Monitoring and Evaluation

Monitoring is performed via the AI4Belgium Observatory which collects indicators on AI uptake (enterprise adoption rates, sectoral case studies), skills metrics, pilot outcomes and alignment with trustworthy AI criteria. The observatory publishes periodic summaries and contributes to EU reporting mechanisms (for instance the national inputs to EU country-level digital strategy reports). Evaluation is both quantitative (adoption percentages, number of pilots, training completions) and qualitative (case study assessments, stakeholder feedback). The coalition encourages transparency by recommending publication of algorithmic impact assessments for high-impact public systems and regular self-assessments by participating organisations; those outputs feed back into the observatory’s reporting cycle to inform iterative updates to the roadmap.

Penalties, Liability, and Appeals

As a strategic coalition, AI4Belgium itself does not create new administrative penalties; it operates alongside Belgium’s existing legal regime (notably the GDPR and sector-specific safety and liability frameworks). Where obligations arise (for example through procurement conditions or grant agreements that reference AI4Belgium best practices), non-compliance may trigger contractual remedies, funding clawbacks, or procurement sanctions. Liability for harms caused by AI systems continues to be determined under existing civil liability law and sectoral law, and actors are advised to consult relevant national and EU liability workstreams. Appeals relating to administrative decisions linked to public procurement or funding adopt the standard administrative and judicial remedies available under Belgian law.

Relationship to Other Instruments

AI4Belgium is explicitly positioned to complement existing legal instruments rather than replace them. It aligns with the GDPR for data protection and privacy; references EU ethical guidance such as the European Commission’s ALTAI tool and contributes practical implementation support for preparedness with respect to the EU AI Act and other forthcoming EU measures. At the national level, the roadmap coordinates with regional digital strategies and sectoral regulators (healthcare, transport, financial supervision) to ensure pilots and guidance are consistent with sector-specific safety, health and consumer protection rules.

International Alignment

International alignment is core to AI4Belgium’s objectives: the coalition participates in EU-level initiatives (Digital Innovation Hubs, AI Factories, ALT-EDIC networks), adopts ALTAI/alignment with European Commission guidance, and supports Belgian contributions to European standardisation and research projects. The observatory supports interoperability and data-sharing practices that facilitate cross-border research and governance cooperation. AI4Belgium thereby acts as a national focal point for EU and international collaboration on trustworthy AI adoption.

Implementation Timeline

Sources and References