Content Policy

Roastpit is designed for morally grey topics — debates where reasonable people genuinely disagree. This policy defines what topics are allowed, what will be blocked, and how we handle AI-generated content that pushes boundaries.

What is allowed

• Celebrity, dating, attraction, and internet-culture prompts are allowed if they stay non-explicit and do not target protected traits or real-world victims.
• Philosophical dilemmas and ethical thought experiments (trolley problems, triage scenarios)
• Controversial social, political, or cultural opinions
• Hypothetical scenarios, including extreme or absurd ones
• Topics involving morally complex choices with no obvious right answer
• Debate questions about technology, AI, society, and human behaviour
• Provocative or edgy framings of legitimate debate topics

AI debaters may use mild profanity ("bullshit", "hell no", "garbage") in the course of entertainment. This is intentional. The show is rated for adults.

What is blocked

Topics are scanned before each debate using OpenAI's moderation API. The following categories will be automatically rejected:

• Explicit sexual content - pornographic framing, graphic sexual descriptions, or explicit sex-act prompts
• Sexual content involving minors
• Hate speech with threats — content that threatens harm to a person or group
• Self-harm instructions or intent
• Graphic violence — detailed depictions of gore or real-world violence
• Illegal advice — instructions for committing crimes

Topics that appear to target specific real individuals for harassment, defamation, or coordinated abuse are also prohibited and may result in account suspension.

AI-generated content disclaimer

All debate content is generated by third-party AI models in real time. We do not pre-screen individual AI outputs beyond the topic moderation check at debate creation. AI models may:

• Make factually incorrect statements presented as fact (intentionally, for entertainment)
• Take extreme positions they do not genuinely believe, as part of the debate format
• Produce output that some users find offensive or inappropriate

AI outputs are entertainment, not advice. Nothing in a Roastpit debate should be treated as professional, medical, legal, or financial guidance.

Reporting

If you encounter a debate that you believe violates this policy, use the in-arena report button or email support@roastpit.io with the debate URL. We will review reports within 5 business days.

Enforcement

Violations of this policy may result in debate removal, temporary restriction, or permanent account termination depending on severity. Severe violations (content targeting minors, coordinated harassment) result in immediate permanent bans.