Page 23 - GSTL_16 April 2020_Vol 35_Part 3

2020 ] FROM DATA MINING TO ARTIFICIAL INTELLIGENCE J43

FROM DATA MINING TO ARTIFICIAL INTELLIGENCE
FROM DATA MINING TO ARTIFICIAL
INTELLIGENCE — WAY TO GOOD
GOVERNANCE
FROM DATA MINING TO ARTIFICIAL INTELLIGENCE
By

Manoj Shrivastav, LL.M.
SUPERINTENDENT, NACIN

When one picks up one’s phone, a number of
push notifications like breaking news alert from a news
portal, game update, response from a dating site etc. are a
common sight. The scope of Artificial Intelligence in day-to-day human life is on
the rise. If a person plans a trip to holy cave of Maa Vaishno Devi after getting
his train reservation through the online booking for the permit slip (yatra parchi
in common parlance) on the portal of the Shrine Board, he would be astonished
to see the welcome sign with all his particulars along with that of the accompa-
nying members. This is work of Artificial Intelligence which has a potential to
determine the risk parameters in GST. Our direct taxation department is effec-
tively using the information generated through this medium but the issue is if the
same could be utilised in GST compliance and augmentation and if yes then how.
Data is the starting point and the end deliverable of every tax task. If the
department does not seize the challenge to manage their tax data effectively, it
will miss the early start incentives. A smarter, faster and more efficient system
using data analytic tools to obtain, analyze and assess underpaid tax and duty
amounts that once took from three months to two years to complete can now be
done on a data-driven basis in a matter of weeks. The churning of data culmi-
nates into an effective Artificial Intelligence based solutions.
Data Analytics tools are being used by tax departments since 2016. The
tool is deployed to find the average tax collection to determine targets for coming
years. For example, if 10 manufacturers of a particular commodity and of a simi-
lar turnover pay ` 10 crore tax, it is unlikely that one company of the same reve-
nue size would pay only ` 1 crore and thus the industry specific targets are ascer-
tained. Second the data analytics could easily point out such anomalies and the
lens would then be put on such units. According to a report published in Econom-

ic Times the department has issued notices to about 200 companies after data
mining revealed that they may be guilty of evading the Goods and Services Tax
(GST).
The findings showed that these companies may be under-invoicing or
selling their goods in cash to their customers. The data mining has raised red
flags in situations where details in GSTR-1 which shows the outward supply,
invoice details and value etc. and GSTR-3B which contains summary of outward
supplies, ITC claimed, and net tax payable didn’t match even in absence of
________________________________________________________________________
 The views expressed in this article are the personal views of the author.
 https://economictimes.indiatimes.com/blogs/et-editorials/mine-data-to-boost-direct-tax-collec-
tions/April 11, 2019.
GST LAW TIMES  16th April 2020  143