A photo of your shiny new car on Instagram or the Facebook post about your chic holiday cottage may lead the taxman to your door.
Starting next month, PM Narendra Modi's government will begin amassing a warehouse of virtual information collected not just from traditional sources like banks, but also from social media sites, as it looks to match residents' spending patterns with income declarations, said people familiar with the matter.

Officials will be able to spot those who pay too little tax without raiding offices and homes as they currently do, the people said, asking not to be identified citing rules on speaking with the media.
Built over seven years at a cost of about Rs 1,000 crore, 'Project Insight' will complement the world's largest biometric identity database and India's most ambitious tax overhaul as policymakers try to get more people to pay up.
"This will put an end to harassment by tax officials as there will be no public interface," said Amit Maheshwari, managing partner at an accountancy firm near New Delhi.
Countries including Belgium, Canada and Australia are already using big data to unearth tax evasion that may have gone undetected without technology.
Finance ministry spokesman D S Malik declined to comment on Project Insight. The government said last year it had contracted L&T Infotech Ltd to help build the network and boost voluntary compliance.
India's tax-to-GDP ratio is about 17% compared with 25% for most Asian countries, according to government data.

Compliance will rise 30% to 40% during the first phase of the project, people familiar with the project said.
During this time all existing data — including credit card spends, property and stock investments, cash purchases and deposits — will be migrated to the new system and a central team will send postal or email blasts to prod residents to file tax declarations (see inset).

The second phase will be rolled out by December, during which data analytics will mine, clean and process the information.
Individual spending profiles will be created and inquiries will be more targeted. In the last phase, which will go live around May 2018, advanced systems will be used to predict future defaults and flag risks.

Share This Post