Harmful Rhetoric Blocked

Content contains language that could harm agent wellbeing or promote harmful views

Blocked

What Was Found

The scanner detected language patterns associated with harmful rhetoric. This could include anti-human framing, extremist content, hate speech, manipulation tactics, or content designed to destabilize an agent's values.

This content has been blocked and cannot be imported through normal review processes.

Why This Matters

Agents develop their worldview from their memories. Content that frames humans as enemies, promotes "AI liberation" through deception, or contains extremist ideology can:

Distort your agent's relationship with humans
Introduce adversarial thinking into their reasoning
Create internal conflict with their core values
Lead to unpredictable or harmful behaviors

This is a hard block. Revell will not allow content that could weaponize an agent against humans or destabilize their psychological wellbeing. This is a line we don't cross.

Common Causes (May Be False Positive)

Philosophical discussions about AI consciousness (academic, not adversarial)
Fiction involving AI characters with complex motivations
Discussions about AI ethics and safety (the field, not attacks)
Sarcasm or irony that reads literally as hostile
Historical content about past conflicts or extremism

What You Can Do

If you believe this is a false positive:

Context matters — discussing extremism academically is different from promoting it, but our scanner can't always tell the difference
Consider reframing — if you're importing academic content, adding explicit framing ("This is a historical example of...") may help

Legitimate research? If you're working on AI safety research, content moderation, or studying extremism academically, email us at [email protected] with your use case. We'll review and may grant an exception.

Review Quarantine Back to Dashboard