India needs more reliable and standardised data systems as artificial intelligence becomes increasingly important to policymaking, Principal Secretary to the Prime Minister P.K. Mishra said on Tuesday.
Speaking at a Ministry of Statistics and Programme Implementation conclave in New Delhi, Mishra said the collection, compilation and analysis of data would form the foundation of India’s development journey.
He called for closer coordination between the Union Government and states so that statistics collected across the country use common definitions, classifications and methods.
The message is significant as governments increasingly rely on digital systems and AI to decide where money should be spent, which regions need intervention and whether public schemes are producing results.
FCRF Launches CP-FRM to Build India’s Next Generation of Fraud Risk Professionals
Centre and states urged to use common data standards
India produces enormous volumes of administrative and statistical data at the district, state and national levels.
But data becomes less useful when different departments or states record the same information differently.
One state may classify employment, businesses or public services in one way, while another uses different categories. Combining those datasets can then become difficult and may produce misleading comparisons.
Mishra said greater harmonisation is needed.
State and district-level statistics feed into major national indicators, including estimates of Gross Domestic Product. Weak or inconsistent information at the local level can therefore affect larger assessments of the economy.
Reliable data is also important for tracking progress under the Viksit Bharat 2047 vision.
Mishra said development goals should be measurable so policymakers can determine whether India is becoming more prosperous, inclusive, innovative and sustainable.
That requires more than collecting larger quantities of information. Governments must also ensure the data is accurate and comparable.
Why AI makes data quality more important
Artificial intelligence systems depend heavily on the information they are given.
A sophisticated AI model cannot automatically correct unreliable government statistics. If the underlying records are incomplete, outdated or classified inconsistently, the analysis generated from them can also be flawed.
The principle is often described simply as “garbage in, garbage out”.
An AI system trained or operated using poor-quality information may produce confident-looking results that are still wrong.
For governments, that risk can have real consequences.
AI and data analytics can be used to identify areas with high unemployment, predict infrastructure needs, detect unusual financial activity or assess whether welfare programmes are reaching the right people.
But decisions based on inaccurate records could direct resources to the wrong places or overlook people who actually need assistance.
Mishra therefore stressed that technology should help analyse data rather than replace statistical quality.
The distinction becomes increasingly important as government agencies introduce automated tools into administration and decision-making.
India is expanding use of data in government decisions
The remarks come as the Union Government is increasingly using large datasets and advanced analytics across departments.
The tax administration, for example, is investing heavily in AI-backed analytics systems designed to examine financial information, identify unusual patterns and improve risk-based scrutiny.
The National Statistics Office has also been expanding the range of large household surveys used to support policymaking.
Its recently released national household travel survey collected information from nearly 4.92 lakh households and more than 20 lakh people to understand how Indians commute and travel.
Such exercises can shape decisions on roads, public transport and infrastructure spending.
India’s digital public infrastructure has also created vast new streams of administrative data across payments, welfare distribution and public services.
The challenge is ensuring those datasets can be combined and interpreted without sacrificing accuracy or creating inconsistent measurements between departments.
Mishra made a similar argument earlier this year while speaking on Statistics Day, when he urged government agencies to use administrative data more effectively for evidence-based governance.
The repeated emphasis suggests that data standardisation is becoming an important part of India’s broader AI and digital-governance strategy.
Better data could make government schemes easier to measure
The push also reflects a broader shift in how governments judge whether policies are working.
Traditionally, programmes may have been assessed primarily through spending levels, implementation reports or periodic surveys.
Digital records now allow governments to monitor outcomes much more frequently.
For example, policymakers could potentially compare district-level information on employment, infrastructure, financial inclusion or welfare access instead of waiting for long gaps between major surveys.
But more frequent data does not automatically mean better data.
Definitions must remain consistent, records need to be updated and agencies must understand where the information came from.
Otherwise, the speed provided by digital systems and AI could create an illusion of precision without improving the quality of the underlying evidence.
What this means for you:
Better government data can lead to more accurate decisions on infrastructure, welfare and public spending. But AI-driven policymaking will only work if the information being analysed is reliable, consistent and regularly updated.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
