The earliest interaction any of us had with Artificial Intelligence was while editing a Word document. This was in February 1971. Now, nearly 47 years later, Artificial Intelligence and Data Analytics surround us. Today, spell checks and grammar are only the most basic applications of machine learning in our daily lives.
Artificial Intelligence and Data Analytics are being employed for a wide range of activities. From consumer analytics for strategically placed ads to personal assistants like Siri, Cortana and Alexa (the latest tool in vogue), AI can handle everything. AI promises us highly capable robots so productive that humans won’t have to work anymore. And it seems, the Indian government’s tax department isn’t far behind in adopting Artificial Intelligence and Data Analytics.
In its 26th meeting, GST Council highlighted a number of major data gaps between the self-declared liability in Form GSTR-1 and Form GSTR-3B.
For the first three months, the GST revenue was over ₹90,000 crore. In recent months, it has hovered around ₹86,000 after a dip. And with the help of data analytics, the Union Ministry of Finance has found the reason behind this leakage in GST collection.
“The theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making and translation between languages.”
As evident, AI is a very broad term comprising a variety of components.
AI encompasses not only cognitive and machine learning but also robotic learning. Robotic learning is when programmers teach machines to perform and imitate steps a human would take to perform a task. It was a Eureka moment when the geniuses realized that they can implement machine learning to perform dangerous or repetitive tasks.
On a daily basis, businesses can use AI for automatic invoice scanning and processing (although this could lead to loss of job for some).
In the tax world, the possibilities of Artificial Intelligence application are endless.
Do the chances of an audit increase when the company headquarters are set up to a new jurisdiction? Does changing the business’s legal structure invite scrutiny? What if you could calculate the probability of audit based on potential sales and expenses for the coming year? Do the possible outcomes of a local or national election impact your business’s tax risk?
AI systems designed for this specific purpose can analyze internal tax data. It can compare the data with an extensive set of data (structured or unstructured) and build a probabilistic model out of it. AI would make this type of analysis possible—even commonplace.
Your legal team would spend hours to go through a 500-page contract to identify possible legal problems or taxation risks. An AI tool can do the same in a matter of seconds. A human fraud analyst works based on experience and intuition and knows what to look for. On the other hand, an AI model can work based on historical frauds and identify patterns by constantly scanning through sales and purchase registers to pick errors.
Major accounting firms have begun applying machine learning and AI to automate book-keeping. Integrating AI for monotonous tasks like data entry and reconciliations removes room for error. The accuracy and speed AI promises could therefore lead to accountants shutting shop.
Artificial Intelligence and Data analytics can be taught to perform any and every function under the sun.
During the Industrial Revolution, machines were invented to undertake activities considered dangerous for humans. Similarly, through automation of advanced technology, AI can take over repetitive human tasks, helping us become smarter, safer society.
Data analytics is an extremely helpful tool in social and financial profiling of taxpayers.
Based on data analytics, system establishes relationships between different entities or people. The data analytics system analyzes different sets of data such as addresses, phone calls, travel trends, I-T returns and now, social media interactions too.
Post Demonetization (which was a trigger of sorts) the government now wishes to analyze both structured and unstructured data. For this purpose, Indian tax officials are advocating Project Insight.
Project Insight is a data analytics platform where algorithms will match residents’ spending patterns (shown on social media) with their declared income. As per a report by NDTV, the government planned to launch the second phase of Project Insight by December 2017.
Expected to go live around May 2018, Project Insight’s analytics enable authorities to predict future defaults and flag risks. Data mining, collection, collation and processing of available information aids the taxmen to monitor high value transactions.
Employing the same logic, the CBEC and GSTN have found inconsistencies between the amount of IGST and Compensation cess paid by importers at customs ports and input tax credit of the same claimed in GSTR-3B.