My Shopping Cart Knew My Salary Before My Mother Did

Dispatch from Neo-Bengaluru, Financial Year 2047. A digital archivist unearths the week of July 28–August 1, 2026, when e-commerce platforms became banks, BNPL became surveillance, and AI underwriting became the world’s most profitable autocracy.

📰 Future Headlines From This Week

• “Flipkart Launches ‘Pay Later’ — Because the ₹299 kurta You Added to Your Cart Also Qualifies You for a ₹50,000 Credit Line” — Moneycontrol, July 30, 2026
• “UAE Credit Bureau Adds BNPL Data to Credit Reports — Your Tabby Installments for Those Slippers Are Now a Permanent Financial Record” — Gulf News, July 2026
• “NPCI Orders Phone Number Masking on UPI — After Only Six Years of Treating Your Mobile Number Like a Public Utility” — Business Standard, July 29, 2026
• “Bajaj Finance AI Bots Originate ₹6,371 Crore in Loans via Chatbots and Voice AI — 27 Autonomous Agents, Zero Human Empathy” — Bajaj Finserv Annual Report FY26
• “OppFi’s ‘Rent-a-Bank’ Scheme Charges 159.79% APR — But They Call It ‘Financial Inclusion’” — NCLC/FDIC Complaint, July 2026
• “Australian BNPL Users Average A$1,110 Outstanding Balance — One-Third Say They Can’t Manage Finances Without It” — Finder Research, March 2026

A Citizen’s Testimonial

My name is Prerana-7, and I am a recovering Flipkart Pay Later user.

In the old world — back in 2026 — when you wanted to buy something, you paid for it. Cash, card, UPI. The transaction was simple: money leaves, goods arrive, relationship ends. Nobody asked about your “purchase patterns” or your “platform behaviour.” The cashier at Big Bazaar didn’t cross-reference your grocery history with your browsing habits before deciding whether you qualified for a credit line.

Then came July 30, 2026. Flipkart, in partnership with PayU Finance, launched “Flipkart Pay Later.” The press release was a masterwork of corporate doublespeak. They called it “putting a decade of commerce signals to work.” They said it would “build an underwriting model that sees these customers clearly.” They described it not as a payment feature, but as “an important step in building a financial services capability.”

Let me translate from the original corporate Sanskrit: We have been watching everything you buy for ten years, and now we’re going to use that data to decide whether you deserve credit.

The product offered three options: Pay Later (30 days), Pay in 3, and EMI plans. Across Flipkart, Myntra, and Flipkart Minutes. Groceries. Fashion. Furniture. Everything. The underwriting model used “commerce intelligence, transaction history, purchase patterns, and platform behaviour.” In other words: that time you bought a ₹999 air fryer at 3 AM during a panic sale? That’s now a data point in your creditworthiness assessment. Your midnight impulse purchase isn’t just a mistake — it’s a financial footprint.

PayU Finance would act as the licensed lender. Flipkart Finance would manage “customer relationships and credit intelligence.” They planned to add more financial institution partners as the platform “expands.”

In my time, we have a word for an e-commerce company that leverages a decade of shopping data to become a lending institution. That word is predator. In 2026, they called it “financial inclusion.”

The BNPL Surveillance State

Meanwhile, the UAE had its own contribution to consumer dystopia this week. The Etihad Credit Bureau announced that Buy Now, Pay Later data from Tabby and Tamara would now appear on credit reports. Starting July 2026. Historical data included.

The framing was predictably noble: “broaden the scope of financial data,” “more inclusive and resilient financial ecosystem,” “bringing more first-time borrowers into the formal credit ecosystem.” The credit bureau said BNPL’s growing adoption made reporting “critical.”

Here’s what was actually critical: a generation of young consumers who used BNPL for everyday purchases — shoes, skincare, takeout — would now have those micro-transactions permanently recorded in a system designed to evaluate their creditworthiness. Every delayed Tabby payment for a pair of sneakers. Every Tamara instalment for a birthday gift. All feeding into an algorithm that would eventually decide whether you get a car loan, a mortgage, a business line of credit.

The irony was surgical. BNPL was marketed as the anti-credit-card — simple, transparent, no interest, no hassle. “Just split the payment!” they said. What they didn’t say: “And we’ll record every single split in your permanent financial file, which will follow you for years.”

In Australia, the consequences were already visible. Finder’s Consumer Sentiment Tracker found 41% of Australians had used BNPL in the previous six months. One-third said they could not effectively manage their finances without it. The average outstanding balance: A$1,110. This isn’t convenience. This is dependence dressed up as a feature.

UPI Privacy: Better Late Than Never (Actually, Just Late)

India’s NPCI, in a move that could only be described as “the bare minimum, six years late,” ordered UPI apps and banks to mask phone numbers. Only the last four digits visible. Username-based UPI IDs would become the default. Deadline: September 2026.

Let’s contextualize this. UPI processes nearly half the world’s real-time retail payment transactions. It has been operational since 2016. For ten years, your phone number — arguably the single most sensitive identifier in the Indian digital ecosystem — was visible to anyone you transacted with. QR code payments displayed it. Merchant receipts showed it. Every chaiwala you ever paid knew your number.

NPCI’s directive cited alignment with the Digital Personal Data Protection Act. Which came into effect in November 2025. Which was passed in August 2023. So the privacy law took two years to implement, and the payment system took another ten months to even begin thinking about compliance.

Experts noted — as experts do, always after the damage is done — that “UPI privacy concerns run deeper than phone numbers.” Storyboard18 reported broader concerns about identity, trust, and data protection in India’s payments infrastructure. But at least now your phone number will be partially hidden. Progress.

The AI Loan Officer Who Never Sleeps (or Thinks)

And then there was Bajaj Finance. India’s largest NBFC. Market cap: ₹6.3 lakh crore. FY26 net profit: ₹16,779 crore. And — here’s the fun part — 27 autonomous AI chatbots that originated ₹6,371 crore in loans during the year. An additional ₹697 crore came from AI-converted voice logs.

Let me put that in language a human might understand: robots that talk and robots that listen decided whether to give real people real money, and they gave out six thousand three hundred and seventy-one crore rupees doing it.

The company’s annual report spoke of this with the enthusiasm of a parent announcing their child’s first steps. “AI is no longer a set of pilots. It is reshaping Bajaj Finance’s business model.” They planned to scale from 27 AI agents to 600+ in FY2027 — across sales, operations, debt management, HR, IT, and risk.

The company even acknowledged — in a filing to the US SEC, because irony is dead — that AI in credit underwriting “may give rise to risks of algorithmic bias, which could result in discriminatory lending outcomes or the inadvertent exclusion of creditworthy borrowers.” Then they continued deploying 27 more bots anyway.

They wanted to build a “next-generation Consumer AI platform.” Offer products on “AI platforms such as ChatGPT enabling chat-based loan origination.” Enable “multimodal AI experiences.” Create a “secure data exchange framework to power AI.”

Translation: We’re going to let an LLM look at your financial life, make a decision in 200 milliseconds, and there’s no human in the loop you can appeal to when it gets it wrong.

Consumer Testimonials From the Archives

“I bought groceries on Flipkart Pay Later. Three weeks later, I got a ‘pre-approved’ personal loan for ₹2 lakh. I didn’t ask for it. They just… knew.” — Meera K., Bengaluru, 2026

“My Tabby payments for Eid gifts showed up on my credit report. The bank rejected my car loan application. I was ‘over-leveraged on micro-credit facilities.’ I bought my sister a scarf.” — Omar F., Dubai, 2026

“The Bajaj Finance chatbot approved my loan in 45 seconds. When I tried to ask it a question about the processing fee, it sent me a pre-recorded message about ’exciting cashback offers.’” — Rajesh P., Jaipur, 2026

“I UPI’d a stranger for a Craigslist purchase. They had my full phone number for two years. They still text me ‘Happy Birthday’ every year.” — Anonymous, Mumbai, 2026

🚨 The Real Threat Behind the Joke

What Actually Happened This Week (Factual Summary)

Flipkart Pay Later Launch (July 30): Flipkart Group and PayU Finance launched an embedded credit product across Flipkart, Myntra, and Flipkart Minutes. Three repayment options: 30-day Pay Later, Pay-in-3, and 3–12 month EMIs. PayU Finance is the licensed lending partner; Flipkart Finance manages credit intelligence using “commerce intelligence, transaction history, purchase patterns, and platform behaviour.”¹

UAE BNPL Credit Reporting (July 2026): Etihad Credit Bureau began incorporating BNPL data from Tabby and Tamara into UAE credit reports, including historical data. The bureau cited financial ecosystem resilience as the rationale.²

NPCI Phone Number Masking (July 29): NPCI directed UPI apps and banks to mask phone numbers across customer-facing interfaces, showing only the last four digits. Username-based UPI IDs to become the default. Full compliance deadline: September 9, 2026. The move aligns with India’s DPDP Act, which came into effect in November 2025.³

Bajaj Finance AI Lending (FY26): Bajaj Finance deployed 27 autonomous AI chatbots that facilitated ₹6,371 crore in loan disbursals. AI-processed 52 million voice logs enabled an additional ₹697 crore. The company plans to scale to 600+ AI agents in FY2027. Their US SEC filing acknowledged risks of “algorithmic bias” in AI-driven credit underwriting.⁴

OppFi Predatory Lending Complaints: Consumer complaints filed with the FDIC documented OppFi’s “rent-a-bank” scheme charging APRs of 159.79% in states with 36% interest rate caps, using national bank partnerships to circumvent state lending laws.⁵

Australian BNPL Dependence: Finder’s Consumer Sentiment Tracker (March 2026) found 41% of Australians used BNPL in the prior six months, with 33% unable to manage finances without it, and average outstanding balances of A$1,110.⁶

What You Can Actually Do About It

1. Opt out of commerce-based credit scoring. If a platform offers you “pre-approved” credit based on your shopping history, decline it. Every acceptance trains the model that your purchase data is fair game for financial decisions.

2. Track your BNPL obligations like real debt. Because they are. Check your credit reports regularly. If UAE-style BNPL reporting reaches India — and it will — every micro-installment will count. Set up a simple spreadsheet or use a free tracking tool.

3. Request UPI ID privacy now. Switch to a username-based UPI ID (e.g., yourname@upi) instead of a phone-number-based one. You don’t need to wait for NPCI’s September deadline.

4. Question AI lending decisions. If an AI chatbot approves or denies you credit, ask for a human review. You have the right. Under RBI’s digital lending guidelines, loan rejection reasons must be communicated. An AI agent’s “decision” is not a legal black box — yet.

5. Report predatory lending. If a fintech lender charges APRs above your state’s cap, file complaints with the RBI (for India) or relevant regulator. OppFi’s 159% APR was exposed because consumers filed complaints. Silence is the fintech predator’s best friend.

6. Read the fine print on BNPL. “Pay in 3” is not free money. Late fees, credit reporting, and the normalization of micro-debt are the real product. You’re not the customer — you’re the product being reported on.

“The cart knows what you earn. The algorithm knows what you’ll borrow. And the chatbot that approved your loan doesn’t know what empathy means. Welcome to 2026.”

— Cashless Consumer