(4/10) Fake Profiles- How Fake Profiles Are Reshaping What We See, Believe, and Trust
Digital Safety Series: Ep. 4
NOTE: All news sources are linked at the bottom of the post under references.
Here’s a number that should stop you: 1,409 accounts. 500 on Facebook, 904 on X, working together, hiding in plain sight for three straight years, posting content designed to look like real people talking.
I want you to sit with that for a second. A coordinated machine, not one person, not one bad actor, and it didn’t need to be clever to work. It just needed to be loud, and consistent, and the algorithm did the rest.
That’s the world we’re actually in now. “Watch out for catfishing” doesn’t even begin to cover it. Let’s talk about what’s really going on. So you know exactly what you’re looking at the next time something online feels a little too perfect, or a little too coordinated.
First Thing First: It’s Not Just One Kind of Fake
Fake profiles exist on a spectrum, and their intentions differentiates what you’re actually up against.
The lonely one: Some fake profiles are almost innocent. Someone builds a better version of themselves online because their real life isn’t giving them what they need - attention, validation, a place to feel seen. Or they’re exploring a part of themselves they’re not ready to attach their real name to yet. This isn’t who we need to worry about.
Catfishing: The one you probably already know - the romance scam. Someone spends weeks building a connection with you, learning what you care about, mirroring it back. Then a crisis hits, and because you feel like you know this person, you send the money. Except they never existed. In India, this has grown into organized scam operations, some running blackmail schemes built around private photos and threats.
The local impersonator: This is the one that shows up in your neighborhood group, your community page. It used to be easy to catch - wrong name, wrong face, something off. Not anymore. Now the names sound like your names. The locations sound like your streets. They comment like they actually live there. The tell is rarely what they say - it’s what they choose to comment on, and how it doesn’t quite fit the way a real local person behaves.
The mob-starter: This is where it stops being about one person and starts being dangerous for everyone. A small group of coordinated fake accounts picks up an unverified rumor, spreads it before anyone officially has confirmed anything - they push it hard and fast across social media platforms like WhatsApp, X. India has lived through this more than once, and it’s not an abstract risk. We’ll get into exactly how in a minute.
The illusion machine: This is where you need one new word: astroturfing. It just means making a fake movement look like a real, grassroots one - like fake grass instead of real grass. There’s nothing organic about it, but it’s built to look organic. A single coordinated network can do this at a scale that’s hard to imagine until you see the actual numbers.
The state-level operation: At the very top of this spectrum sits something bigger than any single account or network, a complete infrastructure. Entire fake media ecosystems, built and maintained for over a decade, designed to shape how whole countries see each other. The goal at this level isn’t to convince you of one lie. It’s to bury you in so many conflicting, fabricated stories that you stop trying to figure out what’s true at all. There’s a term for that: epistemic nihilism. It just means giving up on the idea that truth is even findable. And a society that stops believing truth is findable can’t agree on anything. Can’t trust its own government. Can’t trust its own neighbors.
Six categories, using the same tool but with different intentions. Here’s what it actually looks like when each of these plays out for real.
Why This Isn’t Just Talk
The illusion machine, in practice:
Remember the 1,409-account network from the start of this?
That network, according to NewsGuard, was an India-aligned network. What does that mean? Basically, every post pushed the same direction: Pro-India, Pro-Modi government, Pro-military. It hammered India’s rivals, mainly Pakistan, along with China, the Maldives, and Bangladesh. Investigators never proved exactly who was behind it, they found patterns that pointed toward a possible link to the Indian Army, but they were careful not to claim more than that. I’m not going to pretend it’s more settled than it is.
What is settled: it ran completely undetected from September 2021 through late 2024. It used AI to write and post content, and each account was pushing out up to 10 posts a day. At peak, that’s roughly 14,000 posts a day, flooding the feed you scroll through every morning. Over three years, that adds up to millions of AI-written posts, all designed to look like real people talking.
Around the anniversary of Jammu and Kashmir losing its special status in 2019, the network did something even more targeted. Dozens of fake accounts took on local, Muslim-sounding names - Khan, Bhat and flooded the region’s hashtags with praise about peace and unity. Not to convince anyone of anything specific. Just to drown out the real, dissenting local voices under a wave of fake ones, until the loudest opinion in the room wasn’t a real opinion at all.
The same network also reposted content from real, mainstream Indian outlets and accounts tied to the military, over and over, swarming comment sections until it looked like the whole country agreed on something. In June 2024, a page pretending to be a real news outlet posted criticism of Pakistan’s military. Within minutes, not hours, minutes - 429 fake Facebook accounts inside this same network reshared that exact post, word for word, all using the same hashtag. That’s not a story catching on. That’s a machine flipping a switch, and the platform’s own algorithm mistaking it for real momentum and pushing it to more people.
Investigators ran the network’s text through AI-detection tools, and it came back “highly likely” fully machine-written. A few people, sitting behind some keyboards, could run thousands of convincing fake personas at once. That’s not a conspiracy theory. That’s what generative AI makes possible now.
The mob-starter, in practice:
Before ethnic violence broke out between the Kuki and Meitei communities in Manipur on May 3, 2023, tensions were already being inflamed online. Once the violence actually started, it got worse. A graphic photo started circulating on WhatsApp, Facebook, and X, with coordinated posts falsely claiming it showed a local Meitei woman who had been seriously harmed by Kukis. Fact-checkers traced the photo, it had nothing to do with Manipur at all, it was months old and from a completely different place. But by the time that came out, the image had already done its job. It had already made people want revenge. Similarly, a false claim about a village being burned, ended up leading to actual villages being burned.
When researchers later pulled 2.76 million tweets about Manipur from that first month of violence, they found something telling: only about ~11% of those posts were original. The rest were reshares, retweets stacked on retweets, the same amplification pattern you’d expect from a coordinated network, not from a country having an organic conversation.
This same pattern shows up domestically, too, outside conflict zones. In February 2020, riots in North East Delhi killed dozens of people, fueled in part by rumors and fake reports spreading faster than police could knock them down. Then the recurring “Baccha Chor,” or child-lifter, rumors that have swept through state after state for years, that have led to real lynchings, real deaths, of people who were just passing through.
The state-level operation, in practice:
A 15-year operation called “Indian Chronicles,” was exposed by the watchdog group EU DisinfoLab. It ran over 265 fake local news outlets across 65 countries and resurrected NGOs that no longer existed. It built entire fake identities from nothing, with the aim of moving public opinion.
This is the range of spectrum we are looking at, on one-end 1 lonely person creating a fake profile and on the other end a 15 year operation spread over +50 countries. Taking a fake profile seriously, and reporting it the moment you spot one, isn’t an overreaction. It’s the only thing standing between a rumor and what that rumor can do once enough people believe it. A fake photo takes only a few hours to spread and years to undo the damage it can cause.
How You Actually Catch One
Here’s what to check, before you accept requests, reply to messages, or believe viral posts from any account.
If it’s one account:
Right-click the photo. Run it through Google Lens or TinEye. Real photos rarely trace back to a stock site or someone else’s Instagram. Fake ones usually do.
Check when the account was made versus when it actually started posting. A lot of scammers buy an old, dead account, wipe it, and dump a dozen photos on it in one afternoon to make it look established. Real accounts grow slowly. Fake ones burst.
Look at the follower-to-like ratio. Twenty thousand followers and one or two likes a post? That’s a bought audience.
Check the handle against the name. Someone selling or hijacking an account often forgets to change the handle. “Ananya Sharma” posting as @rahul_verma98 is a dead giveaway.
Ask for thirty seconds on a video call. If the camera’s always broken, or there’s always a convenient emergency, they’re hiding their face because they’re hiding who they are.
If it’s a network:
Watch for the exact same phrase or hashtag showing up across dozens of unrelated accounts within minutes. Real people don’t say the same thing at the same time by accident.
Watch for accounts posting nonstop, every hour, with no breaks. Humans sleep. Bots don’t.
Watch for accounts where every single post attacks the same group or pushes the same one narrative. Real people post about their actual lives too, their food, their family, a bad day. An account that only ever agitates isn’t a person. It’s a tool.
Platforms Should Do More — And So Should We
Ideally, the tech platforms should be doing a lot more than they are. A network of 1,409 accounts ran undetected on Facebook and X for three years, a disinformation operation resurrected dead people and fake NGOs for fifteen years before anyone caught it. That’s a failure of the systems built to catch this at scale. These are companies with the engineering resources and the data to do better. Platforms have a real responsibility here, and they should be held to it, publicly, and often.
Having said that, here’s the part that doesn’t let the rest of us off the hook: even the best moderation system in the world can’t watch every corner of a platform with a billion-plus users. Not in real time, not at the pace fake networks move now. A group with thousands of members can’t have a handful of people checking every profile by hand either. Keeping a community safe isn’t a job you can fully hand off to a company or an overwhelmed admin, not at this scale. It has to be shared.
Reporting a fake profile is genuinely simple. Tap the three dots on the profile whether it’s FaceBook or Instagram or Whatsapp or X, whichever platform it may be. Hit report. Pick the option that’s actually true - fake profile, impersonation. This is basic upkeeping on a space you share with other people, to address something that the platform’s own systems missed.
Nothing about this fixes itself because we hope it will. It gets better because people keep checking, keep reporting, keep asking questions. Verify before you trust. Watch the pattern, not just the words. And when you spot something that isn’t real, don’t just scroll past it. Say something, and use the report button.
References
The India-aligned network (1,409 accounts)
NewsGuard, “NewsGuard Uncovers Massive India-Aligned Network Using AI and Fake Accounts to Target Country’s Foes Operating without Detection for Three Years” (Sept. 2024)
Dark Reading, “Indian Army Propaganda Spread by 1.4K AI-Powered Social Media Accounts”
2020 Delhi riots (misinformation-fueled violence)
Associated Press (via Inquirer), “India’s riot toll rises to 46 as capital remains on edge”
Baccha Chor / child-lifter lynchings
Al Jazeera, “India arrests 18 after two men lynched over WhatsApp rumours”
CBS News, “Police arrest 23 over rumor-fueled lynchings in India”
The Manipur case
The Tribune, “How rumours, fake news fuelled violence in Manipur” (mislabeled Churachandpur video, “Rumour Free” hotline)
BOOM, “No, photo does not show...” (fake photo debunk)
The Print, “No one wants to talk about..” (fake photo, broader context)
Al Jazeera, “In India’s strife-torn Manipur, narrative battle is fought on social media” (2.76 million tweet analysis, 11% original-post finding)
Indian Chronicles (265 fake outlets, 15-year operation)
EU DisinfoLab, “Indian Chronicles: deep dive into a 15-year operation targeting the EU and UN to serve Indian interests”
CBC News, “Huge pro-India fake news network includes Canadian sites, links to Canadian think tanks”
Blackmail scam operations in India