Meta confirms it has deleted 10 million Facebook accounts, here’s why

Meta

Meta has purged around 10 million impersonator accounts from Facebook in the first half of 2025 as part of its renewed crackdown on what it calls “spammy content.” The move is part of a wider clean-up effort aimed at making feeds more relevant and authentic — especially as AI-generated junk floods social platforms.

These fake accounts were reportedly mimicking large content creators to exploit Facebook’s algorithm and audience reach. Alongside them, another 500,000 accounts were penalised for inauthentic behaviour. These accounts were indulging in comment spam, bot-like engagement, and content recycling.

In a blog post, Meta detailed new policies designed to reward original creators, particularly those producing unique images or videos. The company now uses tech to detect duplicate content and suppress its reach unless the original is credited — a direct counter to AI-assisted content farms re-uploading popular material for quick views.

This clean-up effort is happening just as Meta is going all-in on AI infrastructure. CEO Mark Zuckerberg confirmed plans to spend “hundreds of billions of dollars” to build out the company’s supercomputing capabilities and launch its first AI supercluster next year.

As AI tools make it easier — and cheaper — to generate massive volumes of passable content, platforms are scrambling to separate genuine creativity from AI slop. Meta’s rivals are responding too.

Just this month, YouTube updated its monetisation policy, cracking down on mass-produced, repetitive videos that offer little value. The change triggered a wave of confusion online, prompting YouTube to clarify: AI-assisted content is still welcome, but lazy, low-effort clones aren’t.

Meta’s move to delete millions of impersonator accounts — and demote the spam — may only be the beginning of a long-term arms race against algorithm-gaming, AI-fed content mills.