Leveraging Data Innovations for Addressing Information Gaps on SDGs-Fredrik Eriksson.pdf

Explore how data innovations, including the role of AI applications, can contribute to addressing the measurement gaps in the SDGs and well-being, while preserving trust, data sovereignty, and the capacity of national statistical systems to continue providing statistics that adhere to official standards. More information at: https://www.oecd.org/en/events/2025/10/data-innovations-and-initiatives-the-case-for-sdgs-and-well-being

1. AI and Data Innovations at ITU
OECD-WISE webinar: Leveraging Data Innovations for Addressing Information Gaps
on SDGs
9 October2025
Fredrik Eriksson
Senior Data Scientist
International Telecommunication Union (ITU)

2. Data on Internet use/access is scarce –especially in Africa
+ No data collected at sub-national level

3. Track 1: Mobile Phone Data (MPD)
Call Detail
Records (CDR)
Passive
Signalling Data
Active Signalling
Data
Global
Positioning
System (GPS)
Data
Mobile Money &
Transaction
Data
Mobile Network Operator (MNO) Data
App &
Aggregator Data
CRM / Know
Your Customer
Data
Mobility insights
(traffic, tourism, migration,
disaster displacement)
Call Detail
Records (CDR)
Expenditure insights
Activity insights
(information society,
social analysis)

4. Strengths
•Data exists -> high penetration globally
•High-quality data
•Near real-time possibilities
•High geographic granularity
Limitations
•Precision limited by cell tower density
•Representativeness of phone
subscriptions, incl. multiple subscriptions
•Privacy and accessibility –need
agreement with mobile operators
•Large volumes of data -> robust data
infrastructure required
Strengths and limitations of mobile phone data
Actual position
(not visible)
“Observed”
position (visible)
Position of the
mobile device
Position of a
“nearby” cell tower
Network coverage
of the cell (usually
modelled)

5. 10 years of exploring mobile phone data at ITU
Country
projects
International
co-operation
Methodology
and guidelines
Capacity
building
Technical tools

6. Purpose:
•Provide estimates at sub-national level
to support regional and local policy
making
•Improve country-level estimates for
countries without data
•Support targeted digital inclusion
interventions
Track 2: Small Area Estimation using big data

7. Kenya census data
on Internet use at
admin 5 level
Population
density
Distance to ICT
infrastructure
Earth
observations
(e.g.radiance)
100+
covariates
explored
Top-down estimation using random forest

8. Promising signs to use big data to estimate ICT indicators
Tanzania
Big data estimate: 29%
Official ITU estimate: 29%
Malawi
Big data estimate: 20%
Official ITU estimate: 18%
Nigeria:
Big data estimate: 36%
Official ITU estimate: 39%

9. AI to improve access to methodology and data
•AI chatbot to make data and
methodology easier to access
and digest
•Generative “BI” to create
custom charts on demand
•Beta version to be released
next year on

10. Thank you very much!
For any questions, please reach out at:
[email protected]