Large language models are being used to spread spam, but also to fight it. Reddit has developed tools using these models to reduce spam on its platform, blocking 23 million spam views daily and catching 25,000 new spam posts and comments. The company claims a 20% reduction in user exposure to spam from January to March. This is significant because it shows how AI can be used to combat its own misuse, and why it matters is that effective AI-powered moderation could have a major impact on reducing harmful content online.