By Sonny Iroche
Few statistics are more important to the governance and development of a nation than an accurate population figure. A country that does not know, with reasonable certainty, how many people it has, where they live, their ages, economic circumstances and basic social characteristics is effectively attempting to plan its future in the dark.
For Nigeria, this challenge has become particularly urgent.
Nigeria’s last national population census was two decades ago, conducted in 2006. Since then, government institutions, businesses, development agencies and researchers have had to depend considerably on demographic projections and surveys. UNFPA’s 2025 estimate placed Nigeria’s population at approximately 237.5 million people. But projections, however sophisticated, cannot indefinitely substitute for a comprehensive national population and housing census.
Nigeria therefore needs more than another headcount. It needs perhaps the most technologically advanced, transparent and independently verifiable census in its history.
Modern Artificial Intelligence, satellite technology, Geographic Information Systems, cloud and edge computing, encrypted mobile devices, data analytics and carefully governed administrative-data integration now make that possible.
The need for an authentic census figure for Nigeria and any other emerging market cannot be overemphasized as
Population figures affect almost every important decision made by government.
How many primary schools should be built in Kano, Lagos, Abia, Rivers or any other State? How many teachers should be recruited? Where will Nigeria require additional hospitals over the next ten years? How many maternity facilities, doctors, nurses, houses, roads, electricity connections and water systems should be planned? And for me as a former Executive Director in Nigeria’s electricity Transmission Company of Nigeria; a reliable census is so crucial to the planning in our National Electricity Supply Network System.
Government cannot answer these questions reliably without knowing both the size and geographical distribution of the population.
An accurate census is equally essential for determining demographic characteristics: age distribution, household structure, migration patterns, disability, employment characteristics, housing conditions and the differences between rapidly expanding urban settlements and declining or ageing communities.
It also has profound implications for political representation and public-resource allocation. Population statistics inevitably become politically sensitive where communities believe population size may affect representation, constituency arrangements, government programmes or access to public resources.
This makes credibility as important as the final number itself. Nigeria should therefore aim not merely to announce a population figure, but to create a census process sufficiently transparent and scientifically rigorous that Nigerians across political, ethnic, regional and religious divides can reasonably trust its outcome.
International Standards Must Come First
The guiding framework should be the United Nations Principles and Recommendations for Population and Housing Censuses, Revision 4, endorsed by the UN Statistical Commission for the 2030 census round.
Among the fundamental characteristics of a census are individual enumeration, universality within a defined territory, simultaneity, defined periodicity and the production of information capable of supporting detailed geographical analysis.
UNFPA, which supports census programmes internationally, has increasingly encouraged countries to combine traditional enumeration with new technologies, Geographic Information Systems, administrative records, surveys and geospatial information to improve census quality and usefulness.
Nigeria should embrace modern AI tools and technologies in the conduct of future elections, which have become urgent. But AI must assist census professionals, not replace them.
Creating a Digital Map of Every Nigerian Community has also become imperative.
The first stage should begin long before census night.
Nigeria should build a comprehensive National Census Geospatial Base Map covering every enumeration area in the country.
High-resolution satellite imagery, aerial photography where appropriate, GIS information and machine-learning systems can be used to detect buildings, roads, settlements and changes in land use.
Computer-vision models could compare previous maps with recent satellite imagery and identify newly developed communities, informal settlements and peri-urban growth.
This is particularly important in rapidly expanding cities such as Lagos, Abuja, Kano, Port Harcourt, and Umuahia, where entire residential areas may emerge between conventional mapping exercises.
Every enumeration area should have a digital geographical identity. Before a single enumerator begins counting residents, census authorities should therefore have a reasonable estimate of the number of structures and settlements expected within each enumeration zone.
An AI-Assisted Digital Enumeration System comes in very handy here.
Enumerators should be equipped with secured tablets containing encrypted census applications.
Every interview could carry appropriate operational metadata such as the assigned enumeration area, timestamp and authorised device identification, with carefully designed location verification.
The system could prevent an enumerator assigned to one area from casually submitting large numbers of questionnaires apparently originating from another.
AI can then continuously analyse incoming field data.
Suppose an enumeration area normally capable of containing perhaps 600 households suddenly reports several thousand households. The system should flag the anomaly.
If an enumerator supposedly completes an implausible number of detailed household interviews within an hour, the system should flag it.
If hundreds of records contain identical demographic characteristics, repeated household patterns or geographically impossible movements, anomaly-detection algorithms can identify them for supervisory review.
The critical principle is that AI flags; humans investigate and decide. No algorithm should secretly remove citizens from the census or manufacture additions to population totals.
Preventing Double Counting and Under-Counting is another important issue. Perhaps the greatest technical challenge in any census is ensuring that people are neither omitted nor counted more than once.
Nigeria can deploy sophisticated privacy-preserving record-linkage systems to search for probable duplicates based on combinations of demographic and household information.
Where permitted by Nigerian law and strict statistical-confidentiality safeguards, census quality assurance could also draw upon selected administrative information from civil registration, migration, identity and other government systems, not to automatically determine the population figure, but to identify anomalies requiring investigation.
This approach is consistent with the international movement towards integrating census information with surveys, administrative sources and other datasets.
Such systems must, however, be designed with extreme caution. Possession of a national identity number should never become a prerequisite for being counted. A census measures the population according to its defined methodology; it is not simply an aggregation of identity cards.
Finding the People Most Likely to Be Missed should be one of the key guidelines. A credible Nigerian census must pay particular attention to people who are difficult to enumerate: internally displaced persons, residents of informal settlements, nomadic and mobile populations, fishing communities, persons experiencing homelessness, border communities and people living in areas affected by insecurity.
UNFPA’s humanitarian reporting continues to document displacement and insecurity in parts of Nigeria, demonstrating why a conventional door-to-door strategy alone may be insufficient in some locations.
Geospatial intelligence can identify inhabited settlements. Satellite imagery can detect changes in settlement patterns. Carefully designed statistical models can highlight locations in which actual enumeration appears inconsistent with independently observed settlement characteristics.
Again, these technologies should identify where humans need to investigate, not silently manufacture census returns.
The Post-Enumeration Survey: The Census Must Audit Itself
One of the most important components should take place after the initial enumeration.
Nigeria should conduct an independent Post-Enumeration Survey, using a scientifically selected sample of enumeration areas.
Separate teams should return to selected communities and conduct an independent count. Statisticians can then compare those results with census records to estimate:
* people missed;
* people counted more than once;
* households misclassified;
* geographical misallocation; and
* other coverage or content errors.
Nigeria should publish these error measurements. A credible census authority should not claim that its result is “100 per cent perfect.” No population census of a country of hundreds of millions of people can responsibly promise absolute perfection.
The correct objective is a negligible, scientifically measurable and internationally defensible level of error.
Cybersecurity Must Be Treated as National Security
A digital census introduces another major responsibility: cybersecurity.
The census database would constitute one of the largest concentrations of sensitive population information in Nigeria.
It must therefore be protected through encryption in transit and at rest, zero-trust access controls, multi-factor authentication, hardware security, continuous monitoring, independent penetration testing, rigorous incident-response procedures and segregated backup systems.
AI-powered cybersecurity tools can monitor networks continuously for unusual traffic, compromised enumerator devices, unauthorized database access or attempts at large-scale manipulation.
No single official should possess the technical authority to alter final population totals without creating an immutable and reviewable audit trail.
Nigeria AI-Enabled Census Architecture must be close to perfection as much as possible.
Nigeria could structure the system around seven interconnected layers:
Layer 1: Geospatial Intelligence: Satellite imagery, GIS mapping, enumeration-area boundaries, settlement detection and unique geographical referencing.
Layer 2: Digital Field Enumeration: Encrypted tablets, secure census applications, location verification, timestamps, multilingual questionnaires and offline capability for communities without reliable connectivity.
Layer 3: AI Quality Assurance: Anomaly detection, duplicate-pattern analysis, workload monitoring, unusual household detection and automated alerts to supervisors.
Layer 4: Data Integration and Reconciliation: Carefully governed use of civil registration, administrative records, surveys and other legally permissible datasets for quality checking.
Layer 5: Independent Verification: Post-enumeration surveys, academic statistical review, independent technical auditing and controlled observation by accredited stakeholders.
Layer 6: Cybersecurity, Privacy and AI Governance: Encryption, zero-trust security, Nigerian data-protection compliance, model validation, human oversight, audit logs and strict restrictions on access to identifiable information.
Layer 7: National Population Intelligence Platform: Following completion, anonymised census information should feed a secure national demographic platform capable of supporting federal, state and local planning, research and responsible public access. UNFPA itself promotes digital platforms that combine quantitative and geospatial census data and supports wider dissemination of appropriately anonymised census information.
Nigeria must move from a Twenty-Year Census gap, to Continuous Demographic Intelligence. The most important transformation should occur after the census.
Nigeria should not spend enormous resources conducting an advanced digital census only to allow the resulting database gradually to become obsolete.
The census should provide the baseline for a National Population Intelligence System. Subject to legislation and privacy protections, subsequent births, deaths, migration statistics, housing developments and carefully selected administrative information could help Nigeria continuously understand demographic change between censuses.
This would not necessarily replace future censuses. It would make planning between census years considerably more intelligent.
The consequences could be transformational.
Government could know where classrooms will be required before schools become overcrowded. Health authorities could identify rapidly growing populations that require new primary-health facilities. Infrastructure planners could anticipate urban expansion. Emergency agencies could estimate populations exposed to floods or other disasters. Businesses could make better investment decisions. Social interventions could be targeted more intelligently.
Population information would become national planning infrastructure.
Technology however smart, Cannot Substitute for Trust
Ultimately, however, Nigeria’s census challenge is not simply technological. The world’s most sophisticated AI system cannot rescue a census whose institutional independence is questioned, whose methodology is concealed or whose results are perceived as politically predetermined.
Responsible-Human-In-The-Loop (RHITL) Technology must therefore be accompanied by governance, and responsibly applied.
Nigeria should establish an independent Census Scientific and Technology Advisory Council comprising respected Nigerian statisticians, AI experts, demographers, data scientists, cybersecurity professionals, universities and relevant international technical partners.
Its mandate should include reviewing methodology, auditing AI systems, testing algorithms for bias, supervising cybersecurity standards and publishing independent technical assessments.
AI models used during the census should also be documented. Government should be able to explain what each significant model does, what information it examines, what its limitations are and where human intervention is mandatory.
Opacity will undermine trust.
Transparency will strengthen it. Counting Every Nigerian Because Every Nigerian Counts. Nigeria now has an opportunity to transform the national census from a periodic and sometimes controversial headcount into one of Africa’s most advanced exercises in demographic intelligence.
UNFPA describes census information as a key source of population data for sustainable development and promotes its use for evidence-based programming and humanitarian response.
For Nigeria, the stakes are even greater. An inaccurate population figure contaminates almost every statistic derived from it. Per-capita income becomes less reliable. Health ratios become distorted. Education requirements may be underestimated. Infrastructure may be wrongly located. Representation becomes contentious. Resource allocation becomes harder to defend.
Artificial Intelligence will not make a census magically perfect.
But intelligently deployed satellite mapping, GIS, secure digital enumeration, machine-learning anomaly detection, privacy-preserving record reconciliation, cybersecurity, statistical modelling and independent post-enumeration verification can bring Nigeria substantially closer than ever before to a national population figure that is credible, transparent, measurable and defensible.
Nigeria needs to know not merely how many Nigerians there are, but where they are, what their communities require and how those needs are changing.
A credible national census is therefore not simply a statistical exercise.
It is the foundation of political representation, equitable development, investment, resource allocation and national planning.
For a country estimated to be more than 200 million people aspiring to become one of the world’s leading economies, planning by population guesswork is no longer acceptable.
Nigeria must count everyone, count everyone once, count everyone in the right place, and use modern technology to prove that it has done so.
This should be one of the topmost priorities of whoever emerges as Nigeria’s president in 2027.
NB: Sonny iroche Founder & CEO
GenAI Learning Concepts Ltd is one of Nigeria’s leading Oxford- trained AI expert is a member of the Technical Working Group of UNESCO AI Readiness Assessment Methodology and a member of Nigeria National AI Strategy Committee.


