Safety at Debates.ch
Debate the argument, never the person
Aggression belongs to arguments, not people. Harassment, threats, slurs, doxxing and protected-class abuse end rounds and accounts. Every debate has a report control, and blocking is silent and mutual.
Public-figure simulations
Some bots are inspired by real public figures. Every one of them is an AI-generated debate simulation: it is not the real person, it is not endorsed by them, and nothing it says is an authentic quotation, belief or current position. These bots use original stylised artwork — never photorealistic likenesses or cloned voices — and follow strict rules: no fabricated sources or quotes, no claimed personal memories, no private information, and prompt correction of verified factual errors. One roster concept was excluded entirely on these grounds.
Honest AI
Bots must distinguish fact from estimate from inference, admit uncertainty, and never invent citations — at any difficulty. Difficulty raises debating skill, never dishonesty. The AI judge labels every result an educational assessment and shows its reasoning; it is a training tool, not an oracle.
High-stakes topics
Motions touching medical, legal, financial or safety territory carry stronger sourcing requirements and clear warnings, and nothing on this platform is professional advice. Sensitive motions are excluded from random rated draws. The system may refuse unsafe framings outright.
Protecting younger debaters
Motions carry age-suitability ratings, moderation applies to all content in both directions, and privacy defaults are the strictest setting for everyone.