India, the first country to be shorted by AI

Original author: TechFlow

Shiv, a 52-year-old Indian engineer, maintains a routine to this day: sending out at least five resumes daily.

This persistence began in April this year. In March, US software giant Oracle laid off 12,000 people in India, and he was among them. After working for the company for 14 years, he thought he would stay until retirement. Now, he still pays 50,000 rupees in monthly rent. His family has lived in the same house for 15 years, and he doesn't want to force them to move. One evening, he found himself inexplicably snapping at his wife.

In an interview with India's *Outlook* magazine, he said: "We built the technology. We learned it, developed it. Then, after using us, they made us leave."

The same round of layoffs also included 25-year-old Priyanka. She woke up that morning preparing to go to the gym, casually checked her email, and found a cold message notifying her of her termination. She has two installment payments to make: one for an iPhone and one for an electric scooter, totaling 20,000 rupees per month. She is burning through her savings just to stay in Bengaluru.

Zooming out, behind Shiv and Priyanka is a national-scale short-selling reckoning of rare magnitude, and the country being shorted is India.

The World's Purest AI Short-Selling Target is in Mumbai

If you were to find a trading target in the global market that most purely represents the narrative of "AI replacing white-collar human workers," the answer lies both on the long side of the Nasdaq and the short side of the Bombay Stock Exchange. The former is Nvidia, and the latter is India's Nifty IT index.

A glance at this index's trajectory in 2026 looks like a verdict being executed point by point.

The Nifty IT index hit an all-time high of 46,089 points on December 13, 2024, only to retrace 43% by the end of June this year.

In the first half of 2026, the index fell about 30%, making it the worst-performing sector in the entire Indian market. During the same period, the broader Nifty 50 index only declined about 9%. TCS, Infosys, Wipro, and LTIMindtree – India's top four IT giants – have each retreated approximately 50% from their peaks, wiping out a combined market cap of about 19.28 trillion rupees (over $200 billion). TCS alone has seen its market cap fall below the 10 trillion rupee threshold.

What's more intriguing is the rhythm of the decline. Almost every major red candle can be linked to a product launch by a US AI company.

On February 4, Anthropic released a new generation of coding tools, claiming it could automate most of the exploration and analysis work involved in legacy system modernization. COBOL system modernization has been a staple business for India's outsourcing industry for decades. News reached Mumbai, triggering a sell-off in the IT sector, which subsequently fell over 15%, erasing 5.08 trillion rupees in market value.

In May, OpenAI announced an investment of over $4 billion to build a "pre-deployment engineer" team that would embed directly with enterprise clients to restructure workflows around AI. The market immediately understood the subtext: high-value consulting, deployment, and transformation projects could bypass Indian service providers in the future. The Nifty IT index promptly fell to its lowest level since May 2023.

In June, Accenture plummeted nearly 18% in a single day, its worst one-day drop since going public. The next day, when Mumbai markets opened, the Nifty IT index fell 6%, and Infosys dropped 8.19% in a single day to a five-year low, wiping out 1.35 trillion rupees in market value. Accenture's client base – Western banks, retailers, and manufacturers – is precisely the same set of clients served by Indian IT companies.

The sentiment on the sell-side is also shifting.

Investment bank Jefferies warned that in the worst-case scenario, Indian IT stock valuations could still have 30% to 65% downside. A report from Citrini Research anticipates that contract cancellations for TCS, Infosys, and Wipro will continue to accelerate through 2027. Local brokerage Nirmal Bang directly downgraded TCS from Buy to Sell, slashing its target price from 3,046 rupees to 1,693 rupees.

Bloomberg data shows that the combined weight of the top five IT companies in the Nifty 50 has fallen below 7.6%, the lowest level since 2002. The capital market has made its judgment with real money: Global investors are systematically shorting a nation's pillar industry.

The Essence of the Indian Model: Wholesaling Junior Engineers to the World

To understand why India is the most severely wounded by the AI era, one must first understand what the Indian IT industry actually sells.

The answer is straightforward: Engineer hours billed by the hour.

The Y2K crisis at the end of the last century gave India its first big break, and over the following three decades, this model grew exponentially. Clients in New York or London, code written in Bengaluru or Hyderabad. For the same work, an Indian engineer's quote is a fraction of an American peer's. Labor arbitrage is the entire secret behind this $283 billion industry.

This model created an unprecedented social class within India. Neeti Sharma, CEO of TeamLease Digital, summed it up neatly for *Outlook*: "The logic is simple. You borrow four or five hundred thousand rupees to get an engineering degree, join TCS, Infosys, or HCLTech, and your life is set."

The story of an engineer named Pooja is a perfect example of this logic: she grew up in a single room in a Kolkata suburb, sharing a bathroom with nearly 70 people in the building. After getting her diploma in 2005, she went to Gurugram to work as a programmer, starting with a monthly salary of 7,056 rupees. Today, she earns 3.5 million rupees annually at a top IT firm.

A joint study by Nasscom and Crisil shows that by 2007, every IT job was generating about four additional jobs in other sectors of the economy – drivers, security guards, chefs, maids... Housing loans as a share of India's GDP rose from 0.6% in 1995 to about 11% today, with 35% concentrated in the southern states housing major IT hubs. The entire real estate market of Bengaluru and Hyderabad was essentially betting on the paychecks of IT white-collar workers.

The problem is that the product sold by this model has a precise name: the repetitive labor of junior and mid-level engineers.

Writing template code, performing manual testing, maintaining legacy systems, handling tickets... And large language models happen to be the perfect substitute for this type of labor. They are a junior engineer with a marginal cost approaching zero, working 24/7, who will never get a visa and will never need one.

India spent thirty years building itself into the world's largest force for "replacing American programmers." Now, it is being brought down by something cheaper: an "AI that replaces Indian programmers."

The dragon-slayer didn't become the dragon; it was swallowed whole by a new one.

A Decade's Worth of Middle-Class Plans, Torn Up in Three Years

The collapse is already accelerating.

TCS announced in July last year that it would lay off 12,000 people, or 2% of its workforce, the largest layoff in the history of India's biggest private employer. A 45-year-old employee in Kolkata told Reuters: "This is devastating news. For someone my age, finding a new job is incredibly difficult."

An even more absurd detail: over 500 job seekers who received TCS offer letters with a start date of July 2025 are still waiting indefinitely to join. Many of them had already resigned from their previous jobs.

Beyond layoffs, the hiring engine has stalled.

India's top five IT companies saw a net headcount reduction of about 7,000 in the fiscal year ending March 2026, compared to a net addition of over 12,000 in the previous year. Over the past five years, these five companies hired an average of about 230,000 people annually; in FY26, that number dropped to just 170,000. TCS has cut its campus hiring plan from an average of 40,000 over the past three years to 25,000.

Gaurav Vasu, founder of market intelligence firm UnearthInsight, estimates that 400,000 to 500,000 IT professionals face layoff risks in the next two to three years, 70% of whom are in the core group with 4 to 12 years of experience.

Fund manager Saurabh Mukherjea has done an even larger calculation: India produces about 3 million engineering graduates annually, of whom about 1.5 million are considered "qualified engineers." Before 2020, almost all of these 1.5 million were absorbed by the IT services industry. Over the past three years, that number has dropped to nearly zero. Meanwhile, the *India Employment Report 2026* from Azim Premji University shows the unemployment rate for graduates aged 15 to 25 is as high as 40%.

The shockwave is now reversing down the path of wealth creation from previous years.

In the first quarter of 2026, residential sales in major Indian cities fell 13% year-on-year. Analysts directly point to IT layoffs as a primary cause. Shared apartments in Bengaluru are suddenly struggling to find tenants, and landlords are pointing fingers at IT companies. Mukherjea also observes a dangerous signal: A large number of people who sense they will be laid off are rushing to apply for personal loans and home mortgages before losing their jobs. Part of the loan growth in India over the past 12 months came from these "doomsday loans."

So, what about leaving India to work in the USA?

Unfortunately, this path is also being welded shut by Washington.

In September 2025, the Trump administration briefly increased the H-1B visa fee from $5,000 to $100,000 – a 20-fold increase. Two months prior, Trump publicly asked Google and Microsoft to "stop hiring in India."

In 2024, Indians took over 200,000 US work visas, with Indian companies accounting for 20% of all H-1B approvals. This channel was once the physical-world extension of the Indian IT model.

Approximately 60% of the Indian IT industry's revenue, nearly $135 billion, comes from the US market. Today, India faces a double whammy. AI has given US companies a technological option for "reshoring services" for the first time, making it unnecessary to send work to Bengaluru. The new visa policies ensure that Indian engineers will also find it very difficult to send themselves to the US.

People can't go out, and work can't come in.

More frighteningly, the great reckoning brought by AI is still ongoing.

India's median age is only 28. For the next two decades, tens of millions of young people will enter the labor market every year.

The demographic dividend is a check with an expiration date. If cashed in, India becomes the next great power. If not realized, the same young people will shift from the asset side to the liability side of the balance sheet.

A single speck of dust from the times, falling on an individual, becomes a mountain. Shiv is still sending out his five resumes daily. The lights in Bengaluru's office towers remain bright. But for the first time, the people inside are seriously thinking: how much longer will those lights stay on, and for whom are they shining?