Can AI Help Moderate In-game Chat?
summary
Claims that AI can screen in‑game chat in real time: model trains from past human moderation and is ready in two weeks to production, integrates via a speech-to-text pipeline for voice, and flags novel cases for human review. Cited metrics: industry study ~2.3 billion players and 74% encounter harassment, platform survey shows ~70% see toxicity, ~32–40% are targeted. Tooling promises severity scoring, automated temporary bans, language/dialect coverage and over 99% accuracy when trained on platform policy and human labels.
AI-written summary · about
chapters (10)
- 0:00AI in in-game chat moderation
- 3:14Toxic content types
- 7:57Manual moderation limits
- 10:50Ethical AI moderation
- 16:33Building AI moderation models
- 20:10Real-time AI integration
- 24:03AI vs rule-based systems
- 26:56Child safety guidelines
- 29:05Voice and player analysis
- 31:35Q&A
related talks
08mTowards Ethical AI Moderation in Multiplayer Games
42mCurated Panel: A Practical Approach to AI Transformation Across Player Support
45mPanel: From Spam to Speech: Policing the Next Generation of "Unwanted Traffic"
35mPanel: All In Or Incremental? What's The Best Way To Integrate AI In Your Production?
16mFighting Harmful Content with AI: Real-World Safety
38mAI in action: How to use practically in iGaming beyond the hype?
45mEthical use of AI: seeking, understanding, and countering the risks
17mSession: AI Powered Community Relations In App Management