Meta removed around 10 million accounts that were posing as popular content creators in the first half of 2025, a whopping increase in the company’s battle against what it refers to as “spammy content.”
The sheer volume of the purge is one of the most extreme steps that a social media giant has taken to combat the increasing issue of AI-spam inundating online platforms.
Meta Fights Fake Accounts and Spam While Spending Big on AI
The company behind Facebook explained that it made the sensational move as part of a larger strategy to make feeds more genuine and relevant to users. Besides removing the fake accounts, Meta also targeted about 500,000 accounts that participated in what the company considered inauthentic activity and spamming methods.
Instead of complete bans, some of these accounts were subjected to less stringent but still effective penalties. Meta downranked their comments and diminished how far their content gets shared around the platform.
This method is designed to hurt spammers where it hurts them the most their means of earning money from their posts by restricting their reach and interaction.
The move follows Meta’s implementation of tighter controls aimed at incentivizing authentic content from legitimate creators. The company has created new technology specifically aimed at identifying copycat videos and automatically blocking their spread when they get reposted without giving credit to the original creator.
“Unoriginal content is when you use someone else’s videos or photos without crediting the creator,” Meta explained in a blog post unveiling the new policies. This definition gives the company clear grounds to act against accounts that reuse content without adding value or crediting.
The timing is especially interesting in the context of Meta’s huge investment in artificial intelligence technology. CEO Mark Zuckerberg revealed on Monday that the company intends to spend “hundreds of billions of dollars” on AI compute infrastructure, and the first of Meta’s superclusters will go live next year.
The seeming contradiction of spending large on AI on the one hand and battling AI-spam on the other underlines the multi-faceted challenges confronting social media companies in the era of artificial intelligence.
YouTube Fights “AI Slop” with New Policy
Meta is not alone in dealing with these challenges. AI tools’ ease of use has made it extremely simple to create bulk content on social media sites, resulting in what the industry now refers to as “AI slop” low-quality, repetitive content that fills up users’ feeds.
YouTube has just entered the battle with its own policy updates. The Google-owned video site announced this month that content that was mass-produced or repetitive in nature would no longer qualify to be shared in earnings through its monetization programs. The updated YouTube policy starts on Tuesday.
The YouTube announcement created initial dismay among creators, with most thinking that it was a blanket ban on all AI-generated content. But the company quickly clarified its stance, confirming that the policy is against unoriginal and spam content and not against real creators who are using AI tools responsibly.
“We want to support creators who leverage AI tools to augment their storytelling, and channels that incorporate AI into their content are still eligible for monetization,” a YouTube spokesman clarified in a blog entry. The main difference seems to be between creators leveraging AI as a tool to augment their work and those that leverage it to produce formulaic, low-quality content.
Balancing Creation and Combatting Spam
These collective efforts by leading platforms are an acknowledgment that the democratization process of content creation with the help of AI tools has its dark side. While the tools can be used to improve genuine creators’ work, they also allow malicious users to spam the platforms on an unprecedented scale.
The difficulty for businesses such as Meta and YouTube is separating beneficial AI-enabled content from malicious spam. As AI technology evolves further, websites will need to naturally create more sophisticated means of content moderation.
For the users, these should ultimately lead to cleaner, more genuine social media spaces. But whether or not these efforts succeed will depend largely on how successfully sites are able to keep ahead of spammers who are constantly evolving their methods for evading new filtering mechanisms.
