AI Spotlights Over 6.4 Lakh Voter Irregularities in Bengaluru

For the first time since the massive voter‑enumeration drive kicked off, the Greater Bengaluru Authority (GBA) has lifted the curtain on a staggering figure – 6,41,766 logical discrepancies have been spotted across its jurisdiction.

Only about 38.8 % of the enumeration forms were digitised by 4 p.m. on Saturday, yet the AI‑powered system flagged more than six‑hundred‑thousand anomalies, including 2.66 lakh that emerged from Bengaluru Urban alone.

What exactly are these “logical discrepancies”? The official report is deliberately vague, but insiders say they range from mismatched names and fathers’ names to outright duplicate entries. The underlying engine is a bespoke artificial‑intelligence tool that springs into action the moment a Booth Level Officer (BLO) uploads a digitised form.

“The software automatically scans every field and raises an alert if something looks off,” a senior GBA official explained to The Hindu. “It’s the same algorithm that runs our Demographic Similar Entries (DSE) module, which hunts down voters who share identical demographic signatures.”

The DSE tool, as it’s commonly known, is designed to catch both exact copies and subtle variations –‑ a name change here, a typo there –‑ that could otherwise slip through the cracks and inflate the roll.

While the GBA’s daily progress bulletin proudly lists the number of flagged anomalies, the state‑wide report still omits this category, leaving a curious gap in transparency. Chief Electoral Officer V. Anbu Kumar has warned that the final tally of anomalies will only become clear once the draft electoral rolls are published.

With an electorate that now tops 1.03 crore, the sheer scale of the mapping exercise is daunting. Roughly 22.25 lakh electors still lack a proper geographic tag, a shortfall that could, according to some political observers, lead to as much as a 12 % reduction in voter numbers if the discrepancies translate into deletions.

Experts caution that the numbers, while alarming, also point to the power of technology in cleaning up a notoriously messy system. “Artificial intelligence isn’t a magic wand, but it does give us a way to systematically hunt down errors that human eyes might miss,” said a data‑analytics professor at a local university.

As the enumeration progresses, the city’s political landscape may shift –‑ especially if a sizeable chunk of the electorate is pruned based on these AI‑identified flags. For now, the GBA continues to refine its digital workflow, hoping that a more accurate roll will lay a sturdier foundation for the upcoming elections.