Reddit Deploys LLMs to Catch 25,000 Daily Spam Posts as AI Floods Its Forums
Updated
Updated · techbrew.com · Jul 7
Reddit Deploys LLMs to Catch 25,000 Daily Spam Posts as AI Floods Its Forums
3 articles · Updated · techbrew.com · Jul 7
Summary
Reddit said its large language models now flag about 25,000 spammy posts or comments each day, a sign of how heavily AI-generated junk has hit the platform.
That push reflects a broader trust problem: Reddit has no sitewide ban on AI content, but it does ban spam, while many subreddits already require disclosure or prohibit AI posts outright.
The stakes extend beyond Reddit’s own threads because its human-written discussions have become valuable to search and AI systems; Google reportedly pays about $60 million a year for access to Reddit data.
Brands are also using AI to seed product mentions on Reddit to influence chatbot and search answers — a tactic known as answer engine optimization — raising doubts about how authentic online recommendations really are.