evidoria

← Back to browse

Good practice

BPS 2026 Economic Census — AI-Assisted Business Classification at National Scale

Indonesia · Jakarta · See the Indonesia profile

BPS (Statistics Indonesia) deployed generative AI to auto-classify 42 million business entities against the KBLI 2025 standard during the 2026 Economic Census, supporting 251,000 field officers across 38 provinces and geotagging all locations for economic planning.

42 million
Business entities to be classified
251,000
Field census officers deployed
38
Provinces covered
514
Districts covered
2.13 Rp trillion (≈US$134 million)
Census budget

Details

Promoter
Badan Pusat Statistik (Statistics Indonesia)
Period
2026
Keywords
statistics, census, economic data, business classification

Context

Statistics Indonesia (BPS) integrated generative AI into the 2026 Economic Census — the country's first since 2016 — to automatically classify 42 million business entities against the Indonesian Standard Business Field Classification (KBLI 2025). Field data collection runs 15 June to 31 August 2026, deploying 251,000 trained census officers across all 38 provinces and 514 districts.

Activities

The AI system assists officers in real-time business-type classification, handling the increasingly complex post-COVID economic landscape where many enterprises operate across multiple sectors. Alongside AI classification, the system applies geotagging to record the coordinates of all 42 million business locations, enabling spatial economic planning and policy targeting. BPS secured ISO 27001:2022 certification for its data centres ahead of the census.

Results

As the census is currently in progress, post-implementation efficiency metrics are not yet available. The system's evidenced contribution so far is enabling 251,000 officers to perform consistent, standardised business classification at a scale manual coding could not achieve, supported by a Rp 2.13 trillion (approximately US$134 million) budget covering IT infrastructure and field operations.

Implementation
Indicative cost
High (€500k–€5M)
Time to results
Medium (1–3 years)
Staffing & skills
251,000 trained census officers across 38 provinces and 514 districts, BPS technical/data centre staff (ISO 27001:2022 certified)

Conditions for success

  • AI assists but does not replace field officers in real-time business classification
  • Geotagging integrated for spatial economic planning
  • ISO 27001:2022-certified data centres secured ahead of the census
  • Dedicated budget (Rp 2.13 trillion) covering IT infrastructure and field operations

Common failure modes

  • Post-implementation efficiency metrics are not yet available as the census is still in progress
  • The specific AI model and training methodology are not publicly disclosed

Where it fits

Governance type
national statistics agency
Scale
national
Income level
lower-middle-income

Data sources

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

Similar practices you may find useful