Why do we believe fakes? Science reveals the psychology of virals

illustration: DALL-E

A fake image of comet 3I/ATLAS once again proved how easily the internet can turn a simple astronomical observation into a sci-fi story. This viral shows the mechanism in pure form: the more prestigious the supposed source appears, the faster belief in the attractive fiction spreads.

Emotions over evidence. How a viral begins

Researchers from Warsaw University of Technology emphasize in the "Analysis of polarization and inoculation mechanisms (PW/UJ/SWPS)" that algorithms favor bold and flashy content. Images from Hubble and ESA show an ordinary comet: a blurry coma and a tail reflecting light. But they lack that visual "hook" that gives meme potential. A viral acts like a lens — it magnifies what is easy to remember, even if false.

To show the difference between an attractive fake and a real observation, it’s worth comparing two images from the "Analysis...":

Image type	Visual features	Reception effect
Authentic telescope photos	Blurry coma, dust tail, no sharp edges	Lower "clickability"
Viral images	Sharpened contours, added "metallic" elements	Increased engagement and false interpretations

So, we see that manipulation doesn`t have to be sophisticated. Just a few visual tricks are enough for viewers to feel they’re seeing "something extraordinary."

Bubbles, bridges, and leaps. How opinions spread

The model described by researchers from PW in the "Analysis of polarization and inoculation mechanisms" works like a map of social dynamics. It shows two camps connected by bridges — channels of real influence. When bridges are weak, groups stay in their own worlds. When they become too strong, a leap occurs: the entire network adopts a sharp, unified message, regardless of its truth.


This pattern is well illustrated by examples from public health disputes. If one side dominates communication, the other responds with pushback. This leads to waves of trends and countertrends.

What fuels polarization:

• aggressive campaigns targeting only one group,
• poor-quality bridges (lack of trusted intermediaries),
• emotional asymmetry — one side admires, the other rejects,
• lack of preventive education about manipulation.

Researchers show that it’s not the presence of different opinions that raises tensions, but how groups are connected.

Inoculation. How psychological "vaccines" work

Part of the report by UJ describes the inoculation mechanism: a brief training that warns recipients about typical manipulation tricks. The analysis includes exercises that reduce the effect of first impressions and help distinguish falsehood from interpretation.


A real example of such a "vaccine" is the campaign for seniors "Grandma, Grandpa… don’t be fooled." It warns about tricks before a scam appears. This simple technique aligns with the report’s findings: reactions are stronger when the recipient knows the pattern.

SWPS shows another method: an educational game where users play the role of the manipulator. After this experience, people become better at spotting fakes. Results from the study suggest, however, that repeated exposure is needed for a lasting effect. Meanwhile, the SWPS and SGGW team studies awareness of propaganda sources. This method only works in certain conditions. Results don’t automatically repeat, showing how complex social influence processes really are.