Low-quality AI content is changing what people trust and share online

In short: A growing flood of low-quality AI-made posts is making it harder to know what is real online, and it is starting to reshape the shared stories people talk about.

What's going on

“AI slop” is a common label for low-quality content made in huge quantities using AI, including articles, images, videos, and memes. Merriam-Webster has even tied its 2025 Word of the Year, “slop,” to this kind of AI-made content.

Researchers and media monitors say the scale is the point. A report described it as noise that can drown out important signals, like trying to hear a friend in a crowded room. A 2025 study cited by The Guardian found that about 21 to 33 percent of a new YouTube user’s feed could be this kind of content.

This matters because it can change the “shared cultural baseline,” meaning the common set of references people use to talk to each other about news and culture. When feeds fill up with content that looks real at first glance, people may not just believe false things. They may start doubting everything. That can push people to trust only sources that match their existing beliefs.

AI-made “events” can also become cultural touchpoints even if they never happened. Researchers pointed to an AI-generated “wedding photo” of actors Zendaya and Tom Holland that many people accepted as real.

What to watch

Platforms are experimenting with rules like labels for AI-made posts, limits on sharing, and limits on earning money from AI content. Watch whether these steps actually reduce confusion, or whether “AI slop” becomes the default background of the internet, changing what people expect, trust, and remember.

Source: NYTimes