An “elaborate” data analytics and machine learning based investigation by the direct and indirect tax departments has found several people and companies who may have avoided paying taxes. All those identified have been sent notices in July.
The tax departments used statistics from the last two financial years to create an “audit checklist” to identify “outliers” based on industry averages to determine further scrutiny, The Economic Times reported.
It has used this tool for the last few years to identify “discrepancies” in the direct and indirect tax filings–a notable development because historically the two departments operate independently and did not share data.
The shift towards data analytics follows the Centre’s push for big data and has led to “deeper analysis, and more tax notices and scrutiny”, insiders told the newspaper.
Sources further told the paper that government departments that “wouldn’t even talk to each other are now being made to share data”.
The move towards big-data tracking came as data mining by the tax department earlier showed that some companies were under-invoicing or selling in cash to evade taxes. The sharing of data thus brings under scrutiny benami and real estate properties under fake names as well.