- Practitioner Contributions to AI Trustworthiness and Upskilling Standards in CEN-CENELEC JTC 21 WG4
Artificial Intelligence
prEN 18229 will become the harmonised route to presumption of conformity for AI Act Articles 12 to 14. Whether SMEs can actually use it depends on details that only deployers notice: retention periods that are bounded, logs that non-specialist staff can read, a mechanism for a small deployer to obtain conformant logs from an off-the-shelf provider, and oversight competence expressed as tasks rather than qualifications. My comments target exactly these points, grounded in documented deployments with SMEs (a 250-person service provider, small hospitality and construction operators). If adopted, they lower the compliance cost of high-risk AI for the enterprise segment least able to hire compliance staff, reduce dependence on large providers, and make European AI standards a market-access asset rather than a barrier. Medium term this supports faster, safer AI adoption in SMEs; long term it keeps presumption of conformity real for small firms, not only for large ones.
- New standard development
-
This short-term fellowship supported by contributions, via the Czech national standardisation body (ČAS/UNMZ), to the work of CEN-CENELEC JTC 21 Artificial Intelligence WG4 Foundational and societal aspects , and more specifically commenting standards prEN 18229-1 AI trustworthiness framework – Part 1: Logging (CEN/CLC/JTC 21/WG 4, project JT021008); secondary: prEN 18229-3 – Part 3: Human Oversight (JT021056).
Title & Organisation Name: Freelance Consultant
Country: Czech Republic
