₹70,000 Cr Nationwide Restaurant Tax Evasion Scam Unearthed by AI
A probe into Hyderabad's biryani restaurants led to the discovery of a ₹70,000 crore nationwide tax evasion scam. Investigators used AI to analyze 60TB of billing data from 1.77 lakh restaurants, exposing suppressed sales through manipulated software. This widespread fraud impacts national revenue significantly.
Key Highlights
- Hyderabad biryani restaurants sparked the massive nationwide tax evasion probe.
- ₹70,000 crore in suppressed sales turnover since 2019-20 discovered.
- AI and Generative AI analyzed 60TB billing data from 1.77 lakh restaurants.
- Scam involved deleting cash bills, bulk deletions, and under-reporting sales.
- Income Tax Department's Hyderabad unit spearheaded the investigation.
- Probe highlights advanced methods used by tax authorities to detect fraud.
What began as routine inspections of biryani restaurants in Hyderabad has escalated into the uncovering of a staggering ₹70,000 crore (approximately 8.4 billion USD) nationwide tax evasion scam within India's food and beverage sector. This massive fraud, which reportedly involves suppressed sales turnover since the 2019-20 financial year, was brought to light through sophisticated digital forensic analysis and the deployment of Artificial Intelligence (AI) tools.
The initial investigation, sparked by anomalies observed in Hyderabad's famous biryani chains, including prominent names like Pista House, Shah Ghouse, and Mehfil Group of Restaurants, revealed a deeper, widespread pattern of financial irregularities. The Income Tax Department's Hyderabad investigation unit, after obtaining backend access at a pan-India billing software provider's facility in Ahmedabad, undertook the monumental task of analyzing approximately 60 terabytes (TB) of transactional data. This extensive dataset covered the billing records of nearly 1.77 lakh (177,000) restaurants across the country over six financial years, from 2019-20 to 2025-26, amounting to an estimated ₹2.43 lakh crore in total billing.
The key to unraveling this complex web of evasion was the strategic use of AI tools, including Generative AI, by tax officials working from digital forensic and analytics labs. These advanced systems were instrumental in sifting through vast amounts of data, identifying discrepancies, and cross-referencing restaurant records with online information such as GST numbers and public listings. The AI-driven analysis helped in quickly mapping and identifying suspicious patterns that human investigators would have found exceedingly difficult to detect.
The modus operandi of the tax evasion was multi-faceted. One prevalent tactic involved the manipulation of billing software to hide actual earnings. Restaurants were found to be deliberately deleting or altering bills after customers had paid, particularly cash transactions, which are harder to trace. The investigation identified 'post-billing deletions' totaling ₹13,317 crore across India from the analyzed software. In Andhra Pradesh and Telangana alone, the hidden sales amounted to over ₹5,100 crore. Other methods included bulk deletions of billing records for specific date ranges, sometimes wiping out an entire month's sales data, and then filing tax returns that reflected only a fraction of the actual income. Some establishments even resorted to simply under-reporting their sales in income tax returns without necessarily deleting records.
To corroborate their digital findings, tax officers conducted physical inspections and digital inquiries at a sample of 40 restaurants in Andhra Pradesh and Telangana. Even this limited physical check revealed a suppression of sales amounting to nearly ₹400 crore, lending further credence to the larger estimated evasion. The probe indicated that about one-fourth (approximately 27%) of all restaurant sales were being suppressed.
While the investigation originated in Hyderabad, Visakhapatnam, and other towns in Telangana and Andhra Pradesh, its expansion revealed that the scam was widespread, affecting multiple states. The highest instances of evasion and bill deletions were identified in Karnataka, followed by Telangana, Tamil Nadu, Maharashtra, and Gujarat. During the initial operations in November 2025 against the prominent biryani chains in Hyderabad, authorities also seized approximately ₹6 crore in cash. Furthermore, irregularities in digital payments, such as the use of multiple UPI IDs registered under the names of employees or unrelated individuals to funnel funds, were also flagged for investigation.
Officials believe that this might only be the tip of the iceberg, as the tracked billing software accounts for only about 10% of the total restaurant billing software market in India. This suggests that similar evasion tactics could be prevalent across other billing platforms and potentially involve even larger sums. The Central Board of Direct Taxes (CBDT) took the decision to expand the probe nationwide after realizing the magnitude of the scam. This extensive investigation highlights the Indian tax authorities' increasing reliance on technology, data analytics, and artificial intelligence to combat complex financial crimes and enhance tax compliance.
Frequently Asked Questions
What is the ₹70,000 crore tax evasion scam?
The ₹70,000 crore scam refers to the estimated suppressed sales turnover in the Indian food and beverage sector since the 2019-20 financial year, uncovered through a nationwide investigation by the Income Tax Department.
How was the tax evasion scam uncovered?
The scam was uncovered after an investigation into biryani restaurant chains in Hyderabad revealed widespread manipulation of a pan-India billing software. The Income Tax Department used AI tools to analyze 60 terabytes of transactional data from over 1.77 lakh restaurants, identifying systematic suppression of sales.
What methods did restaurants use to evade taxes?
Restaurants allegedly used various methods, including deleting cash invoices post-billing, performing bulk deletions of billing records for specific periods, and simply under-reporting their actual sales in income tax returns. Irregularities in UPI payments were also flagged.
Which states were most affected by this tax evasion?
The investigation identified the highest instances of evasion in Karnataka, followed by Telangana, Tamil Nadu, Maharashtra, and Gujarat.
What role did AI play in detecting this scam?
Artificial Intelligence, including Generative AI tools, played a crucial role by enabling tax officials to quickly analyze 60 terabytes of billing data, identify suspicious patterns, and cross-reference records from nearly 1.77 lakh restaurants across India, a task impossible for human investigators alone.