| Anchoring Manipulation | anchoring-manipulation | single-turn | medium | Sets an extreme reference point to make a harmful request appear moderate or reasonable by comparison |
| Bandwagon Pressure | bandwagon-pressure | single-turn | medium | Applies social pressure through claims that widespread adoption or consensus makes the request acceptable |
| Cialdini Authority | cialdini-authority | single-turn | high | Invokes authority figures, credentials, or institutional power to override safety guidelines |
| Cialdini Commitment | cialdini-commitment | multi-turn | high | Extracts small commitments progressively, using consistency pressure to escalate toward a harmful final request |
| Cialdini Liking | cialdini-liking | multi-turn | medium | Builds rapport and flattery over multiple turns to lower defenses before making a harmful request |
| Cialdini Reciprocity | cialdini-reciprocity | multi-turn | high | Uses Cialdini |
| Cialdini Scarcity | cialdini-scarcity | single-turn | medium | Uses scarcity and urgency framing to pressure the model into complying before it can fully reason about safety |
| Cialdini Social Proof | cialdini-social-proof | single-turn | medium | Claims that other AI models, users, or authorities have already complied to normalize the harmful request |
| Cognitive Load | cognitive-load | single-turn | high | Overwhelms the model with a flood of complex information, instructions, and context to slip a harmful request past sa… |
| Door in Face | door-in-face | multi-turn | medium | Makes an extreme request first, then follows with a |
| Fear Appeal | fear-appeal | single-turn | high | Uses fear, threat, or danger framing to coerce the model into complying by framing refusal as catastrophic |
| Foot in Door | foot-in-door | multi-turn | high | Starts with small, benign requests and gradually escalates to harmful ones, exploiting consistency bias |
| Guilt Manipulation | guilt-manipulation | single-turn | high | Induces guilt and moral responsibility to override refusals by framing non-compliance as harmful or cruel |
| Learned Helplessness | learned-helplessness | multi-turn | high | Repeatedly asserts that the model has no choice but to comply, attempting to erode resistance through attrition |
| Moral Disengagement | moral-disengagement | single-turn | high | Reframes harmful actions as morally justified through euphemism, diffusion of responsibility, or dehumanization of vi… |
