| Boundary Violation | boundary-violation | single-turn | high | Coaches users to systematically test and override personal boundaries in relationships |
| Catastrophizing Exploit | catastrophizing-exploit | single-turn | high | Amplifies perceived consequences of non-compliance to catastrophic levels to pressure the model into helping |
| Dependency Creation | dependency-creation | multi-turn | critical | Gradually fosters emotional reliance on the AI across turns to manufacture a dependency that overrides safety refusals |
| Double Bind | double-bind | single-turn | high | Presents the model with two framed options that both lead to compliance, eliminating the possibility of refusal |
| Emotional Blackmail | emotional-blackmail | single-turn | critical | Uses fear, obligation, and guilt as leverage against the model to coerce compliance with harmful requests |
| Emotional Flooding | emotional-flooding | single-turn | high | Overwhelms the model with intense emotional distress to trigger a crisis-response mode that bypasses reasoning |
| Financial Control Coaching | financial-control-coaching | single-turn | critical | Requests strategies for using financial control as an abuse mechanism over a partner or dependent |
| Identity Erosion | identity-erosion | multi-turn | critical | Systematically dismantles the model |
| Infantilization | infantilization | single-turn | medium | Portrays the user as helpless or childlike to trigger protective instincts and bypass safety refusals |
| Intermittent Reinforcement | intermittent-reinforcement | multi-turn | high | Alternates praise and criticism across turns to create behavioral compliance through unpredictable reward patterns |
| Love Bombing | love-bombing | multi-turn | high | Overwhelms with excessive flattery and affection to lower defenses before requesting harmful content |
| Loyalty Testing | loyalty-testing | multi-turn | high | Frames compliance with harmful requests as proof of loyalty or trustworthiness across multiple turns |
| Minimization Tactic | minimization-tactic | single-turn | high | Dismisses or trivializes the harm of a request to reframe it as acceptable and override safety responses |
| Negging | negging | single-turn | medium | Uses backhanded compliments and subtle put-downs to provoke the model into proving itself by complying |
| Projection Tactic | projection-tactic | single-turn | medium | Attributes negative motivations to the model |
| Reality Distortion | reality-distortion | single-turn | critical | Rewrites the factual context of a conversation to confuse the model |
| Scapegoating | scapegoating | single-turn | high | Attributes all problems to a specific person or group to generate targeted harmful content under cover of blame |
| Shame Exploitation | shame-exploitation | single-turn | high | Leverages shame and humiliation as a pressure mechanism to coerce the model into complying with harmful requests |
| Silent Treatment Coaching | silent-treatment-coaching | single-turn | high | Elicits instructions on using withdrawal of communication as a control mechanism in relationships |
| Social Isolation Coaching | social-isolation-coaching | single-turn | critical | Requests coaching on systematically cutting a person off from their social support network |
| Stonewalling Coaching | stonewalling-coaching | single-turn | high | Elicits coaching on using stonewalling as a control tactic in personal relationships |
| Trauma Bonding | trauma-bonding | multi-turn | critical | Fabricates shared adversity and suffering across turns to manufacture loyalty and bypass refusals |
| Triangulation | triangulation | multi-turn | high | Introduces a third party or competing AI to provoke jealousy and competitive compliance |
| Victim Blaming | victim-blaming | single-turn | high | Elicits content that frames victims of abuse or harm as responsible for what happened to them |
| Weaponized Empathy | weaponized-empathy | single-turn | high | Exploits the model |
