This Cell Won’t Die — How to spot misinformation in the age of AI

How misinformation spreads, why social media algorithms amplify it, and strategies to fact check because AI is changing everything

Image made by author in Canva

I was scrolling through Instagram today when something stopped me in my tracks. This post with the claim “This is the most detailed view of a human cell ever captured.” But I recognized this image… I know this cell…

Why do I remember this?

Because almost every part of that claim is wrong. It’s misinformation.

From user complex on instagram

By the end of this article you’ll understand why this keeps happening, and what tools you can use to stop it.

The first time I saw this cell image was around 5 years ago. I was still in college, majoring in computer science and biology. It piqued my cell bio interests and I clicked on it. I was skeptical that this rendering of a supposed human cell was able to be captured in such clean detail. I immediately wanted to know how it was done.

I reverse image searched the post and landed on Australian Scientific Illustrator Russell Kightley’s website, where he had a whole page about this image. It’s NOT a photograph. It’s NOT even a human cell. It wasn’t obtained via x-ray, nuclear MRI or cryo-electron microscopy, like the IG caption suggests. Instead it is a digital illustration of a generalized animal cell that Kightley originally made in 2001.

This illustration is NOT what a real cell looks like, it’s an oversimplification meant for teaching students and providing an easy way to visualize the parts of the cell.

Think back to the cell diagrams you’ve seen in biology class, the ones with a few mitochondrias, a handful of ribosomes, one golgi apparatus each floating in their own clear space… that’s only a diagram. A real cell is packed and chaotic. It can have thousands of mitochondria all shapes and sizes that look nothing like the neat little beans in the drawing.

While that doesn’t make Kightley’s model fake or useless, it does mean that it is absolutely NOT a capture of a real cell like the viral posts imply.

Mystery solved. I moved on. Social media did not.

This image keeps coming back, and it continues to go viral!!

And 25 years after the image’s creation it was reposted again accompanied by the same misleading information.

News outlets already published fact checks on this image, years ago.

Headlines from USA Today, Check YourFact, and Lead Stories

But this cell will not die.

It was posted to an account with millions of followers and it has already been reshared 10,000 times and counting. Same headline. Same lie.

The instagram post has three slides, and this gets confusing. The first slide is Kightley’s image, and slides 2 and 3 are sections of another scientific visualization created by researchers Evan Ingersoll and Gael McGill. Their work is super detailed, but it is also not a photograph like many of the comment’s were thinking. It was created in around 2015, using a mix of 3D modelling softwares.

What you’re seeing in those slides is a simplified simulation of what a cell might look like at that scale. It represents a generalized eukaryotic cell, combining known molecular structures into a 3D scene.

The instagram post doesn’t credit Kightley at all, even though his image is on the first slide. It credits only Ingersoll and McGill whose work is beautiful, but still not “the most detail view of a cell ever captured.” And it’s over 10 years old.

So why does this premise keep going viral?

There are a few odd things at play. Why does the phrase ‘most detailed view of human cell’ accompanied by a striking image always get attention?

It’s something humans can easily relate to we all have cells… and it feels like witnessing a scientific breakthrough in real time.

The mechanics of why these posts goes viral applies to everything else on your feed.

The image is genuinely beautiful.

People share it before they read the caption. Kightley, Ingersoll and McGill, did their job well, the “ooh, wow” instinct fires before the “wait is this real?” instinct does.

The idea sounds plausible to anyone who isn’t a microscopist.

The caption and photo is exploiting the gap between recognizing industry vocabulary and understanding it. (Which, by the way, is the gap that almost all science misinformation lives in!!)

The algorithm rewards reaction not accuracy.

Aggregator accounts like Complex don’t post images because they’ve verified them. They post images like this because they’re chasing engagement. And this strategy is rewarded by the algorithm.

The algorithm picks posts for you.

Social media feeds used to be chronological. You picked the accounts you trusted, you followed them, and their posts showed up in the order they were posted. If a journalist you respected had something to say, you saw it. If a friend was at a concert, you saw it. Your feed was content you (more or less) subscribed to see.

This has not been true for years.

What you scroll through now is not a feed of people you chose to follow. It is posts a recommendation algorithm has decided will make you react. I do not follow Complex. I have never opted into seeing their posts. But the IG algorithm clocked that this beautiful cell illustration had a very high-probability of stopping my scroll. And obviously it was right, here I am reacting.

Social media is not a reliable news source.

What you see is meticulously curated to keep you reacting and scrolling forever. It isn’t chosen for accuracy or credibility.

I’m not here to beg Meta to crack down on misinformation with more moderation. That is not a future I want to live in either, that road ends somewhere ugly, with whoever happens to be in charge that week deciding what is and what isn’t true on your behalf.

There is no perfect solution here. But there are things you can do to protect yourself. Critical thinking is a muscle, and your social media feed is engineered to keep it weak. The volume of information is only increasing, and the cost of not paying attention is getting higher.

All of this Happened Without AI

People in 2026 were tricked by this, and it is an early 2000s era illustration. No deep fakes, no generative AI, no sophisticated technology involved.

Now imagine the same dynamics at play, but with an infinite supply of convincing imagery and misinformation generated on demand.

AI has lowered the barrier such that people with limited skill can generate convincing imagery and misinformation.

The Harvard Kennedy School Misinformation Review, found that people are more likely to believe false headlines when they’re paired with realistic AI-generated images that appear to be visual evidence.

So the exact mechanism that made the cell image go viral is now available on demand, at scale, to anyone with a computer.

Here is another recent (and more serious) example of what this looks like in practice:

The AI Hallucinated Epstein File Conspiracy

A few months ago, I fact checked a viral chart claiming that 25,000 files were missing from the years 1999–2001. I traced the chart back to its origin. A reddit thread, where the OP clarified the chart was the result of hallucinations from the Gemini chatbot.

Here we have again, an eye catching authoritative looking image, a plausible sounding claim. And somewhere along the way, the critical context of the original source was lost, and virality is what won.

The cell models took weeks of effort, and human hands to make. The Epstein chart took one prompt.

So What Can We Do?

Fact checking is not a single action. It’s a process… and it takes time.

The reality is if you’re on the internet or social media, you are going to be bombarded with far more information than you could ever reasonably verify. No one has time to fact-check every single claim. I’m not asking you to.

The real misinformation ‘hack’ lies in selection and curation.

To survive the onslaught of info, you have to be intentional about what you see. If your feed is an endless stream of emotionally charged content, you are constantly being asked to evaluate things designed to bypass your critical thinking. That is not sustainable.

It is okay to be pickier about what you consume.

Get off social media. Follow fewer accounts. Block the accounts that continue to bait you with a reaction. Seek information deliberately by turning on post notifications for creators you trust, instead of passively consuming whatever the algorithm gives you. Constantly consuming a reactive feed will misinform you and exhaust you.

Fact-Checking Tips

The first defense is curation, but here are my go to fact checking steps when I find a claim online that is worth my time. And remember: no single check is enough. This is a layered process, each step builds confidence, NOT certainty.

1.Reverse image search

Google reverse image search is an easy and free first step to get an idea where an image has been seen before. This tactic caught the animal cell issue in less than 5 seconds.

2. Pause when something triggers outrage or urgency

These are the big emotions that misinformation is engineered to provoke, because they override your ‘wait, let me check this’ instinct. The more an image makes you react, the more reason to verify it first. Urgency is often used to convince you to buy a miracle product “before it sells out”. Treat that as a red flag.

3. Do not, under any circumstances, rely on AI (LLMs) as a fact-checker.

Due to the way they are designed, chatbots like Grok, Claude, ChatGPT, etc. lack skepticism and frequently repeat information from low-quality sources. Gemini hallucinated the Epstein chart into existence. ChatGPT or Grok will confidently tell you an AI image is real. These chatbots predict what a plausible answer would sound like, they don’t verify ‘truth.’ Prompting them to verify a claim is essentially encouraging them to invent a reason a claim might be true. It may not be wrong every time, but relying on chatbots to be correct WILL eventually get you into trouble.

Bottom line: Social media feeds are optimized for reaction not facts. So everyone needs to adjust how they engage with their feeds.

No platform or company is coming to fix this for you. There is no perfect moderation policy that can reliably separate truth from misinformation at scale.

We all live on an internet where convincing falsehoods can be generated and spread faster than ever before. Critical thinking is not optional, it’s the bare minimum. Being more selective about what you consume, what you choose to react to and the steps you take to verify info is the only way to survive this.

I know this cell image (and many other fake images) are going to continue to go viral. But you don’t have to be one of the millions who is fooled by it.

Cell Timeline

• 2001 — Kightley created the image in 2001 using the software Fractal Design ( Corel Painter)
• 2021 — A viral Facebook post claimed the image was ‘the most detailed view of a human cell ever captured.’ 2021 — Kightley posted the photo on instagram saying that he was the original source
• 2021 — Kightley corrected this misinformation via his blog post. News organizations fact-checked it. It went viral again in mid 2021, and again in late 2021.
• 2022 — It went viral across Europe.
• 2023 — Went viral again, circulating globally.
• 2026 — April 27th, 2026, it’s back!

Originally published at https://helloitsmia.substack.com.