loading . . . AI-generated image of Pope Francis wearing a white puffer jacket. Source: screenshot.
In March 2023, a photo of Pope Francis wearing a white puffer jacket spread like wildfire across social media. The image looked so convincing that many people believed it was a real photograph. Only later did it emerge that it had been created with the AI image generator Midjourney.
The photo was not propaganda. But it showed the world something far more important. For the first time, the wider public could see with their own eyes how quickly and convincingly artificial intelligence could create content that was almost impossible to distinguish from a genuine photograph.
Today, the concern is no longer limited to individual AI-generated images. What is increasingly alarming is the mass production of such content. When the internet is flooded within a short period of time with thousands of AI-generated images, videos, articles and posts, finding reliable information becomes increasingly difficult.
A new term is being used more and more often to describe this phenomenon: **slopaganda**. Over the past year, the word has become increasingly common in the vocabulary of disinformation researchers and journalists.
## How Slop Became a Propaganda Weapon
The word _slopaganda_ combines two English words: _slop_ and _propaganda_.
Online, _AI slop_ refers to quickly produced, low-value or mass-generated artificial intelligence content. It may be a hastily created Facebook image, yet another version of the same meme, or an automatically generated article whose only purpose is to fill the information space.
When this kind of content is used to spread political narratives or influence public opinion, it becomes slopaganda.
The term itself is still evolving. It is being used increasingly often by journalists, technology analysts and disinformation researchers, although there is not yet a single agreed academic definition. Researchers of digital politics at Columbia University describe slopaganda as a phenomenon in which generative AI is used to mass-produce cheap influence content whose purpose is not so much to persuade as to flood the information environment.
## The More, the Better
While traditional propaganda sought to create a single powerful message, slopaganda follows a different logic. Quality is no longer what matters most. Quantity is.
When the same narrative appears in hundreds of Facebook posts, dozens of videos and countless AI-generated comments, it begins to feel familiar. Psychology has a well-known concept called the illusory truth effect: the more often we encounter a claim, the more likely we are to perceive it as true, even when there is no evidence to support it.
That is why slopaganda does not need to be perfect. There simply needs to be a lot of it.
## States Are Already Using AI
This is no longer a future scenario.
In 2025, OpenAI published a report describing dozens of cases in which its language models had been used in attempts to support state-linked influence operations. According to the report, networks linked to Russia, China, Iran and North Korea used large language models to write social media posts, translate articles, generate comments and adapt propaganda messages for different target audiences.
In one Russia-linked operation known as Stop News, ChatGPT was used to write propaganda scripts, create video prompts and translate content into French. These materials were then used to produce AI-generated videos that were distributed through TikTok accounts targeting African audiences.
Source: screenshot.
In the same report, OpenAI also describes an operation in which ChatGPT was used to create video scripts and image prompts praising Russia’s Africa Corps. These were then used to generate videos with another AI model. The report’s authors note that the French subtitles even contained grammatical errors, suggesting that the text had not been edited by a human after being automatically translated.
Source: screenshot.
At the same time, OpenAI stresses that there is still no evidence that artificial intelligence has made such influence operations significantly more effective. It has, however, made propaganda faster and cheaper to produce, as well as easier to scale.
## **Slopaganda ei tähenda ainult võltsfotosid**
Sageli seostatakse slopagandat AI abil loodud võltsfotodega, kuid tegelikult võib see avalduda väga erineval kujul. See võib tähendada sadade peaaegu identsete kommentaaride avaldamist mõne uudise all, automaatselt loodud veebilehti või lühivideoid, mis kordavad sama sõnumit eri vormides.
Propaganda ei pea olema ilmtingimata realistlik. Propastop on varem kirjutanud ka AI abil loodud LEGO-stiilis sõjavideotest. Need ei püüdnud vaatajat veenda, et nähtav on päris. Nende eesmärk oli muuta sõda mänguliseks, emotsionaalseks ja kergesti jagatavaks. Ka selline sisu võib kuuluda slopaganda alla, kui seda toodetakse ja levitatakse masstoodanguna poliitiliste narratiivide võimendamiseks.
A LEGO-style animation depicts a geopolitical conflict in a playful and simplified form. Source: screenshot.
Propastop has also written about how China-linked influence networks used large language models to adapt propaganda content for audiences in European countries. In these cases, artificial intelligence did not create new narratives. Instead, it made the production of existing narratives many times cheaper and faster.
This kind of automation does not necessarily generate new propaganda narratives, but it allows existing messages to be produced on a much larger scale and adapted for different languages and target audiences.
## Will Slopaganda Replace Propaganda?
Hardly. Rather, it is changing the way propaganda is produced and distributed.
Artificial intelligence does not create new propaganda mechanisms, nor does it automatically make people easier to influence. It does, however, make it possible to produce vast amounts of text, images, videos and comments much faster and at a far lower cost than before. When social media algorithms begin amplifying this content as well, the information environment can fill with AI-generated material faster than people are able to verify it.
Slopaganda therefore represents more than just a new propaganda technique. It describes a transformation of the entire information environment. While influence operations once focused on individual, carefully crafted stories or campaigns, the same goal can now be served by an enormous volume of cheap, repetitive AI-generated content.
The greatest challenge posed by future information operations may no longer be a single convincing lie, but an endless stream of mediocre content that makes reliable information increasingly difficult to find.
Post Views: 48 https://www.propastop.org/en/2026/07/20/propaganda-glossary-slopaganda-how-artificial-intelligence-is-turning-propaganda-into-mass-production/
