₹ 70,000 Crore on the Plate: From Biryani Bills to Big Data in India’s Tax Crackdown
A few months ago in this column, we spoke about how the tax department’s newest star performer is no longer the one with a briefcase full of files, but the one hidden inside a server room. We saw how analytics exposed bogus refunds and suspicious donation patterns, and I had remarked that the future of tax enforcement would depend less on surveys and more on servers.
Well, the servers seem to have been working overtime — and unlike humans, they don’t take tea breaks or forget files.
The latest example comes from what may become one of the biggest technology-driven tax investigations in India’s restaurant sector. What began as a probe into a few Hyderabad biryani outlets has reportedly uncovered a possible ₹70,000 Crore suppression of restaurant sales nationwide — a signal that the enforcement playbook may have changed permanently.
The investigation reportedly focused on a widely used restaurant billing software platform. The system is used by over one lakh restaurants covering around 10% of the market. Instead of checking books shop-by-shop, authorities reportedly analyzed nearly 60 Terabytes of billing data covering 6 financial years. The dataset covered about ₹2.43 lakh crore of transactions across roughly 1.77 lakh restaurant IDs.
What the analysis revealed was striking. It reportedly found patterns suggesting under-reporting of at least ₹70,000 crore in sales. Karnataka, Telangana and Tamil Nadu reportedly showed prominent deletion patterns. In several cases, restaurants generated full internal bills but later removed selected invoices — especially cash transactions — before filing returns, as if deleting from software erased the reality. Unfortunately for them, servers have long memories. And unlike old diaries, you cannot blame the office boy for misplacing them.
Within the larger suppression estimate, nearly ₹13,317 crore was linked to suspected post-billing deletions. Sample studies indicated that in certain cases as much as 27% of actual turnover may not have been reflected in official disclosures. If such ratios were to be extended across a wider base, the potential size of under-reporting in the sector could be far larger — though that remains speculative for now.
To verify the digital findings, authorities reportedly carried out physical and electronic inspections at selected restaurants. These checks revealed significant undisclosed turnover, strengthening the credibility of the data-driven conclusions. Importantly, tax liabilities and penalties are still being computed, meaning this story may continue to unfold in the months ahead.
What makes this episode important is not only the quantum involved but the technique used. The department reportedly used big-data analytics and AI tools, including generative AI, to map GST numbers, match billing patterns and detect anomalies across millions of entries. In short, the investigation did not begin with a notice but with a dataset — a clear shift where discoveries now lead to surveys rather than the other way round.
The larger lesson is that electronic records are no longer just an accounting convenience — there are becoming statutory audit trails. Earlier, manipulation inside internal systems often stayed hidden unless a survey team arrived. Today, centralized data analysis can identify suspicious patterns remotely. Volume, which once concealed discrepancies, now exposes them.
This also shows how the nature of tax evasion itself has changed over time. In earlier years, unrecorded sales were often tracked through kachhi chittis, pocket diaries or informal noting. Once those papers were destroyed, the trail often vanished with them permanently. Today, things are very different. Even if an electronic entry is deleted, it often leaves a digital footprint — in logs, backups or reconciliation mismatches. In the digital era, records may disappear from screens, but they rarely disappear from systems — the cloud remembers what the drawer used to forget.
This also brings us to a structural concern in India’s tax framework. Barely two crore individuals carry the bulk of the income-tax burden in a nation of over 140 crore people. When high-cash sectors suppress turnover, the burden quietly shifts onto compliant taxpayers, salaried employees and organized businesses. In that sense, such investigations are not merely enforcement exercises — they are steps toward balancing the tax system.
There is a policy angle too. If technology can identify irregularities in just 10% of the software market, similar analytics may soon extend to retail, e-commerce, logistics, healthcare and professional platforms. The tax officer’s new field visit may well be to a data centre rather than a business premises.
This trend is also supported by regulatory amendments. Today, companies, whether listed or unlisted, are required to maintain books of account in accounting software with proper audit trails. The intention is clear — financial records are expected to be traceable, verifiable and tamper-evident. As digital reporting systems mature, the scope of such requirements is only likely to widen further across sectors. In other words, technology is turning book-keeping into a verifiable digital history.
Of course, skeptics exist. Some argue that AI-based analysis may sometimes misread normal business corrections as manipulation. Others say cancellations, billing errors and system glitches are routine in the hospitality industry. These are valid points, and due process must always follow digital detection. After all, algorithms can raise suspicion, but only verification can establish liability. Yet the direction is unmistakable. Tax administration in India is moving from notice-driven scrutiny to data-driven detection. Instead of asking questions first and searching later, authorities are increasingly searching first and asking questions later. Going forward, reconciliation across GST, income tax, payment systems and internal software is no longer optional — it is a survival strategy.
The biryani-to-big-data episode may ultimately be remembered less for the ₹70,000-crore figure and more as the moment when technology clearly overtook traditional enforcement methods. In the AI era, the safest tax planning strategy is still honesty — backed by clean systems and consistent reporting. Because today, every invoice leaves a footprint. And unlike before, those footprints are now being tracked not by inspectors with files, but by machines with memory.
Perhaps the biggest takeaway is this: the time has changed. This is not the era to invent smarter evasion techniques; it is the era to adopt smarter compliance practices. When every transaction leaves a trace and every dataset talks to another, the real competitive advantage for businesses may no longer lie in tax avoidance tricks but in clean systems, transparent reporting and compliance-driven growth.
[Views expressed are the personal view of the author. Readers are advised to seek professional advice before taking any decisions. Readers may forward their feedback & queries at nareshjakhotia@gmail.com. Other articles & response to queries are available at www.theTAXtalk.com]
