Latvia's AI Centre Sandbox — First Cohort Selected From 10 Applications, €370,000 State-Funded
Latvia
Latvia's newly created AI Centre ran the country's first AI regulatory sandbox call: 10 applications, 3 projects selected, free access …
Luxembourg · Luxembourg City · See the Luxembourg profile · See the Luxembourg City profile
Top 75% 47/100 · Ask Evidence Copilot about this practice
Luxembourg's data-protection authority CNPD has run 'Sandkëscht,' a rolling AI sandbox, since May 2024. Independent research found 2 public pilots by mid-2025, but the CNPD itself has disclosed no participant counts or exit reports.
The Commission nationale pour la protection des données (CNPD), Luxembourg's data-protection authority, launched "Sandkëscht" on 21 May 2024, opening applications from 14 June 2024, ahead of a formal EU AI Act sandbox requirement; participation runs 9 to 18 months across three phases.
CNPD has not published participant or application counts on its own channels. An independent 2026 study for the Austrian Chamber of Commerce (JOANNEUM RESEARCH) identified two publicly mentioned pilots as of July 2025 — AI-assisted speech analysis for customer interaction, and algorithmic risk models in finance — and noted that the sandbox runs from the CNPD's regular €10.3 million annual budget and 67-strong staff rather than dedicated funding.
Each project is meant to conclude with an internal "exit report" (gap analysis), and CNPD announced a systematic review process from autumn 2025, but as of August 2026 no exit reports or aggregate results have been made public; the sandbox also has no dedicated legal basis beyond general GDPR cooperation powers.
As a caution, with only two pilots identified over more than two years, and no self-published throughput or outcome data, this is a longer-running but lower-disclosure sandbox than Latvia's or Colombia's — a useful contrast case on transparency practice.
Read the full analysis: https://cnpd.public.lu/fr/actualites/national/2024/05/lancement-sandkescht.html
Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Latvia
Latvia's newly created AI Centre ran the country's first AI regulatory sandbox call: 10 applications, 3 projects selected, free access …
Saudi Arabia
SDAIA's Data & Privacy Regulatory Sandbox, launched Feb 2023 as the region's first, lets MSMEs test privacy-enhancing AI and data …
Brazil
Brazil's data authority ANPD opened a regulatory sandbox in August 2025, letting three selected firms test AI systems under direct …
Norway
Since 2021, Norway's data protection authority has run about 20 time-boxed AI sandbox projects with public bodies — from NAV's …
Open full copilot Grounded in cited practices — always check the sources.