Test-Driving Social Media Moderation with AI 'What If' Simulations

Can we test moderation before it goes live?

Researchers built an AI-powered simulator that replays the same online conversation under different moderation rules—like running parallel "what if" worlds—so we can measure how toxic talk changes while keeping everything else equal.

• LLM agents act with striking psychological realism.
• Toxicity spreads via social contagion (one bad post fuels others).
• Personalized moderation outperforms one-size-fits-all policies.

Why it matters: real-world experiments are costly and messy. This approach lets platforms compare strategies safely, quickly, and transparently.

"Same conversation, different rules, measurable outcomes."

Authors: Giacomo Fidone, Lucia Passaro, Riccardo Guidotti. Categories: cs.AI, cs.CY, cs.MA. arXiv: http://arxiv.org/abs/2511.07204v1

Paper: http://arxiv.org/abs/2511.07204v1

Register: https://www.AiFeta.com

AI ContentModeration SocialMedia LLM AgentBasedModeling OnlineSafety ToxicSpeech Research