turning knowledge into a decision advantage

All He Wanted Was a Treat Instead Came Pranks on Repeat

He made the information accessible. The internet made it actionable. That's the difference good KM needs to understand.

Elizabeth Raju

10/14/20264 min read

Your phone number and address are the keys to your front door and this Diwali season, one man discovered exactly what happens when you hand those keys to the entire internet. The same is true of the data we leave lying around at work, especially now that AI can read it.

The Story

An influencer wants a box of sweets for Diwali. So he does what feels friendly and harmless - he posts his home address and phone number for his followers. Maybe a few boxes of sweets arrive. He probably thinks this is the best idea he has ever had. Happy ending, roll credits. Except the internet was not done with him yet. Next comes the real twist. He starts receiving everything you can imagine - a refrigerator, then a washing machine. Then smartphones. Then smart TVs. Screenshots claiming a military tank had been ordered to his address circulated online and even made the news. The laddus were the free sample. The bill came after.

What actually happened?

According to reports, people used his name, number and address to place cash on delivery orders on the various shopping and food apps. COD requires no card and requires no payment upfront so anyone could 'gift' him anything.

  • Strangers placed COD orders in his name, about 15 in half an hour at one point.

  • Delivery guys turned up with goods he had never ordered and asked for cash and OTPs.

  • More people got involved, sharing fake order screenshots purely to annoy him.

  • He posted a video asking people to stop. The story got into the national newspapers and it was all very embarrassing for him.

Then came one more video. He announced that he had left his house, was travelling from place to place, and would not share his whereabouts with anyone, so nobody could send him COD packages any more. Lesson learned: if they cannot find you, they cannot deliver to you.

Funny for everyone except the people in it

Full confession: When I first saw the military tank screenshot ordered from Amazon I laughed too. Then I saw his video requesting people to stop the prank and the reality struck. Every prankster thinks they added one little joke. Together they have built a harassment machine and two groups paid for it.

The influencer: his phone would never stop ringing. Strangers would knock on his door all day long asking for money he didn't owe. In the end he lost his peace, his privacy and for now, his home. That is the real punchline, and nobody is laughing at it.

The delivery workers: they are the unseen victims. They rode across the city in heavy traffic with heavy appliances, got turned away at the door, and rode back with nothing to show for it. Multiply that by dozens of fake orders.

The pranksters spent few seconds and zero money. The workers lost hours, fuel and income. The person at the address lost his peace of mind.

From his doorstep to our desks: company data in the age of AI

He posted his details to the whole internet. At work, we make a quieter version of the same mistake - personal and confidential data left in shared folders, team sites, old email threads and forgotten project files. And now AI can read all of it. Workplace AI assistants do not leak data on purpose. They find and repeat whatever they have access to. If a customer's phone number sits in an old project file shared with "Everyone", an AI assistant can surface it to any colleague who asks the right question.

Data cleansing means fixing errors, duplicates and inconsistencies so that data is accurate. Data redaction means hiding or removing sensitive details, like names, numbers, addresses and ID numbers, before content is shared, published or fed to AI. A black box drawn over text in a PDF is not redaction: the text can often be copied out underneath.

Why this matters even more for knowledge management

  • Oversharing becomes discoverable. A file shared too widely used to stay hidden because nobody went looking. AI goes looking for everyone.

  • Old content is a liability. Outdated drafts, duplicate files and old contact lists (called ROT: redundant, obsolete, trivial) feed AI wrong answers and expose data nobody needs any more. Data you no longer keep cannot be leaked.

  • AI connects the dots. An employee ID in one file and a home address in another were harmless apart. Together, they make a profile, just like his name + number + address did.

  • Public AI tools sit outside the company. Pasting a client list, a contract or a colleague's details into a public chatbot is sharing it outside your organisation.

In KM terms, cleansing and redaction are not just housekeeping. Clean, well-governed content is what makes AI answers safe and trustworthy. Before you store, share or feed anything to AI, ask: does this need to contain anyone's personal details?

Data hygiene checklist, five minutes now saves a crisis later.

  1. [ ] Check who has access before you share. Avoid "Everyone" for anything sensitive.

  2. [ ] Keep personal data out of shared folders unless the work truly needs it.

  3. [ ] Redact personal details before sharing documents externally or uploading them anywhere new.

  4. [ ] Use only company-approved AI tools, and never paste confidential or personal data into public chatbots.

  5. [ ] Clean out ROT content on your sites and drives every quarter: old drafts, duplicates, outdated contact lists.

  6. [ ] Apply sensitivity labels to confidential documents.

  7. [ ] Before you share anyone's details, even casually, ask: would they want this shared?

  8. [ ] Never post your home address or personal number publicly.

If data gets out, act fast and report early. The steps below are general guidance; follow your company's own policy first.

  1. Stop the spread. Do not forward, download or screenshot it further. Note where you found it.

  2. Report it right away to IT. Companies may have legal duties to report breaches, so early reporting matters.

  3. Restrict access if you own the content.

  4. Help clean up: find copies in other folders, chats or email threads.

  5. Turn it into learning. Update the guideline, template or training so the same leak does not happen again. That is knowledge management at work.

The moral of the laddu

Laugh at the story, sure. A man asked for sweets and the internet tried to send him a tank. Absurd! But remember who carried the fridge, and who left his home. Protect your own details, and never be the one who turns someone else's details into a joke.

At work, every overshared folder is a small public post. With AI reading it, the risk is even bigger. Clean it, redact it, and share only what is needed.

Honestly, I feel for him. He simply trusted his followers and never imagined what could happen once his details reached the internet. I wish we could trust without having to think twice about what might go wrong. But the internet has a way of taking something shared in good faith and carrying it far beyond what we ever intended.

Sources: Reported in national media in October 2026, including Business Today, Free Press Journal and News Karnataka.