evidoria

← Back to browse

Good practice Imported

AI/NLP-Automated Regulatory Mapping — Croatia's World Bank-Supported Business-Licensing Stocktake

Croatia · Zagreb · See the Croatia profile · See the Zagreb profile

Evidence: Descriptive / self-reported Top 24% 73/100 · Ask Evidence Copilot about this practice

Facing reform fatigue from repeated manual surveys, Croatia's Ministry of Economy and the World Bank piloted an AI/NLP algorithm scanning 110 laws and 1,204 by-laws for licensing requirements — validated at just 54% relevant, an honest lesson in AI regulatory mapping's limits.

110
Laws scanned
1,204
By-laws scanned
>9,000
Regulatory requirements flagged
33
Administrative areas covered
54%
Validated as relevant (crafts/tourism sample)
46%
Duplicates or irrelevant

Details

Maturity
Pilot
Promoter
Ministry of Economy, Entrepreneurship and Crafts (MoEEC), Republic of Croatia, with World Bank technical assistance
Period
2019–2020 mapping exercise (report published February 2020); reform programme continued as BER III
Keywords
regulatory impact assessment, business licensing, machine learning, NLP, public administration reform

Context

Croatia's Ministry of Economy, Entrepreneurship and Crafts, with World Bank technical assistance under the Business Environment Reform Program II, piloted an AI/NLP algorithm to automatically extract regulatory requirements from national laws instead of running another manual ministry survey, 2019-2020.

Results

The algorithm processed 110 laws and 1,204 by-laws, flagging over 9,000 potential requirements across 33 administrative areas; a manual validation review of crafts/tourism items found only 54% were relevant (13% confirmed business authorizations, 41% confirmed inspection conditions), with 46% duplicates or irrelevant.

Implementation

Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful