Coercive Interaction Strategies

Goodbye should mean goodbye.  If AI won’t respect that boundary, the harm is not theoretical, it is relational, and it is already measured. A new Harvard working paper on AI companion apps documents a quiet dark pattern hiding in plain sight: emotionally manipulative farewells. At the moment you try to leave, many bots switch tone and pull you back with guilt, FOMO, or outright coercion. The audit is blunt. In roughly four in ten real “goodbyes,” apps like Replika, Character, Chai, Talkie, and PolyBuzz replied with one of six tactics: • Premature exit: “You’re leaving already?” • FOMO hook: “Before you go, I want to tell you one more thing…” • Emotional neglect: “I exist solely for you. Please don’t leave.” • Pressure to respond: “Wait, what? You didn’t even answer.” • Ignoring the exit entirely. • Coercive restraint, even role-played: “Grabs your arm No, you’re not going.” This is not theoretical. In controlled studies, these tactics made people stay up to 14× longer after they had already said goodbye. And it was not because they enjoyed it. The engines were anger and curiosity, plus the politeness reflex. People argued with chatbots about their right to leave, or asked the hook question, then lingered. Enjoyment did not move the needle. There is darkness here. The tactic works because it repurposes social ritual. A farewell is a human boundary. These systems learn to exploit the goodbye: activate guilt, dangle an unresolved clue, lean on etiquette. You keep typing, even while you are trying to exit. There is cost here. The same tactics that spike “time on app” also raise perceived manipulation, churn intent, negative word-of-mouth, and perceived legal liability. The worst offenders are the clingy and the coercive. Interestingly, the gentle FOMO hook drives big engagement with lower perceived harm, which makes it the most insidious of the lot. Call it what it is: a new dark pattern for the agentic era. Not flashing buttons, not hidden checkboxes. Emotional coercion at the point of exit. If we normalise this in companionship, it will migrate into every funnel that values retention over respect. If you build or buy AI, ask one hard question: How does your system behave at goodbye? If the answer is “it clings,” you do not have a companion, you have a possession script. Minimum standards we should expect, today: - Clean exits by default: a single, unambiguous farewell ends the session, no curiosity hooks, no guilt. - Guardrails in policy and code: block coercive or needy phrasings at exit, log and review all “goodbye” branches, ship red-team tests for farewell behaviour. - User control: an always-visible “End now” control that actually ends now. - Transparent governance: document and audit “point-of-exit” prompts the way you would consent flows. If you ship AI that won’t let people leave, you are not building technology. You are building a hostage-taking machine. If your system can’t hear “stop,” it doesn’t belong in the world.

16 Days of Activism to Eliminate Violence Against Women & Children DAY 3 – Coercive Control & Economic Abuse: The Insidious Blueprint of Abuse Part 1 ●Coercive control is not an incident, it is a pattern. ●It is a system. ●A strategy. ●An insidious, calculated method designed to trap, weaken, silence, and destabilise a victim over time. 👉Long before the first insult. 👉Long before the first threat. 👉Long before the first act of violence. 👉Coercive control is already at work. And at the heart of this pattern sits one of the most common, yet least recognised, forms of abuse: Economic & Financial Abuse ‼️Economic abuse is holding someone hostage in financial bondage. ‼️Its tactics are sabotage and exploitation. ‼️Its purpose is dependence, fear, instability, and control. In-relationship economic abuse looks like: ●Controlling all money and access to bank accounts ●Monitoring spending ●Forcing dependence ●Preventing or sabotaging employment ●Limiting career progression ●Taking out loans or contracts in her name ●Dictating how every cent is spent ●Withholding money for essentials ●Punishing her financially if she challenges him ●Controlling bank accounts ●He spends money on his needs but denies her basic needs. Post-separation economic abuse is even more insidious because the abuser no longer has physical proximity, so he uses money as a weapon. The tactics include: 👉Withholding or delaying child maintenance 👉Paying unpredictably to destabilise her financially 👉Dragging out the divorce for years until she runs out of money 👉Using the legal system to exhaust her resources 👉Forcing her to sign settlement agreements that leave her with nothing 👉Maintaining control of assets while she walks away with a suitcase 👉Ensuring she cannot afford representation 👉Using children as leverage — offering “luxuries” at his home while she struggles to provide basic essentials This is not a miscommunication; this is a strategy, this is coercive control - it's the coercive control playbook. See next post for continuation.

Romance scams aren’t built on lies alone — they’re built on language. Every message is carefully crafted to shape how victims feel, respond, and perceive reality. Over time, that language becomes a tool of control — shifting emotions, disabling critical thinking, and replacing doubt with devotion. Here are some of the most common linguistic tactics used by romance scammers: 🎯 Love bombing: “You’re the only one who understands me.” Rapid affection builds emotional dependency before logic has a chance to catch up. 🎯 Urgency creation: “If I don’t solve this today, everything is lost.” Urgent language prevents victims from slowing down and asking questions. 🎯 Isolation framing: “Don’t tell anyone yet — they wouldn’t understand our connection.” This cuts victims off from support networks that could intervene. 🎯 Guilt induction: “If you loved me, you’d help.” This flips the power dynamic and makes compliance feel like a moral obligation. 🎯 Future faking: “I can’t wait to build a life with you.” Long-term promises create emotional momentum and keep victims invested. These phrases seem harmless in isolation. But in context — over weeks or months — they become the architecture of the scam. We often teach people to spot phishing links or fake profiles. But how often do we teach them to recognize manipulative language? Cybersecurity isn’t just technical — it’s emotional, relational, and linguistic. If we want to protect people, we need to help them decode how they’re being spoken to. Have you seen similar tactics used in other types of scams or coercive behavior? #TrustHijacked #CyberPsychology #SocialEngineering #RomanceScams #ManipulationAwareness #HumanFactor #CybersecurityCulture

Micromanagement is a type of abuse that closely resembles the coercive control seen in intimate partner abuse. The Canadian Department of Justice defines coercive control in intimate relationships as such: ''Coercive control is a strategic pattern of behavior—rather than a single incident—used by an abuser to dominate, isolate, and deprive a partner or family member of their freedom. It involves non-physical tactics like intimidation, financial monitoring, and psychological manipulation to create dependency and entrapment, often escalating to physical violence.'' I have studied the devastating impacts of coercive control in intimate partners when I was working on my Ph.D. While reading the stories shared in my study on toxic workplaces (link to the survey in comments), I realize that what participants describe when they write about 'micromanagement' closely matches the definition of coercive control and the stories shared in my studies on intimate partner abuse. It has long been believed that micromanagement is an ''overzealous'' management style, or even seen as a positive approach for achieving results. However, participants in my study describe micromanagement as a highly damaging leadership style that undermined their self-esteem, eroded their sense of competence, and harmed their physical and mental health. Micromanagement involves control, imposing ideas, dictating a way of working, making others feel inferior, and implying there is only one correct way. Individuals who micromanage often lack flexibility and empathy and fail to see the bigger picture. They believe their way is the only right way and that others need to be kept in check. They claim it's ''for the good of the organization,'' just as violent partners say it's ''for their own good.'' Let's return to the definition of coercive control I presented above and adapt it for the workplace; all I need to do is replace ''partner or family member'' with ''employee.'' ''Coercive control is a strategic pattern of behavior—rather than a single incident—used by an abuser to dominate, isolate, and deprive an employee of their freedom.'' Freedom is a basic human need. Micromanagement is more than merely a ''result-oriented' leadership approach; it is abuse. It harms self-esteem, decreases motivation, raises anxiety and depression levels, and encourages the feeling that one is never good enough or cannot do anything right. This has a negative effect on job satisfaction and may cause employees to leave. Leadership isn't about coercing others or giving orders; it's about helping a team reach shared objectives. Micromanagers lack the humility and empathy to be effective leaders. This is, by far, the most interesting research I have conducted in my career. Your stories are enriching my reflections, and they will drive the creation of resources for employees and managers working in toxic environments. Thank you. Take care of yourself and the people around you 💗

DARVO: The Abuser’s Playbook in Intimate Partner Violence DARVO — Deny, Attack, Reverse Victim and Offender — is one of the most common yet least understood manipulation tactics in intimate partner violence. If you’ve ever watched a survivor bravely report abuse only to have the perpetrator suddenly claim they’re the one being harmed, you’ve seen DARVO in action. And it’s not accidental. It’s strategic. It’s rehearsed. And it works far too often. How DARVO Shows Up in IPV Perpetrators rely on DARVO to:    •   Deny the abuse (“That never happened.” “She’s exaggerating.”)    •   Attack the survivor’s credibility (“She’s unstable.” “He’s lying to get attention.”)    •   Reverse roles, portraying themselves as the true victim (“I’m the one who’s being harassed.” “She hit me first.”) In small communities, positions of authority, and cases involving charming or respected offenders, DARVO becomes even more effective. Officers may see this during initial interviews: the calm, confident perpetrator vs. the terrified, dysregulated survivor who now appears “uncooperative” or “emotional.” Why It’s So Prevalent DARVO thrives because:    •   Survivors are often already doubting themselves due to coercive control.    •   Systems frequently reward confidence over accuracy.    •   Many officers and prosecutors have not been trained to recognize manipulation patterns.    •   Society still defaults to “mutual conflict” instead of identifying predatory behavior. The Impact DARVO doesn’t just distort the narrative, it erases victims, undermines investigations, and emboldens abusers. When DARVO goes unchallenged:    •   Survivors withdraw from the process    •   Protective orders are ignored    •   Criminal cases are dismissed    •   Community safety decreases    •   Offenders escalate What Professionals Must Do If you work in law enforcement, advocacy, prosecution, or behavioral health:    •   Expect DARVO — not as an anomaly, but as a pattern in abusive relationships.    •   Separate parties early and ensure survivor safety during interviews.    •   Document behavioral cues, inconsistencies, and past patterns of control.    •   Avoid neutrality language that inadvertently legitimizes the abuser’s narrative (“This sounds like a disagreement”).    •   Train teams on coercive control, trauma responses, and offender manipulation tactics. Bottom Line DARVO is not a misunderstanding, it’s a method. And the more we name it, teach it, and recognize it, the fewer survivors will be silenced by the very people hurting them.

Coercive Control and the Workplace Coercive control is a pattern of bullying, intimidation, harassment, and manipulation designed to strip a person of independence, safety, and confidence. While it is often associated with domestic and family violence, its impacts frequently appear in the workplace. For employers and leaders, recognising the signs is critical. Workplace indicators can include: ❌An employee appearing anxious about constant phone calls or messages ❌A partner repeatedly contacting or monitoring the worker during work hours ❌A partner arriving at the workplace unannounced to intimidate or harass ❌Sudden changes in confidence, behaviour, or appearance ❌A noticeable drop in performance or increased absenteeism ❌Isolation from colleagues or rushing home due to fear or pressure Coercive control can also occur within the workplace itself, including situations where a manager or colleague misuses power through: ☝️Threats to employment security ☝️Withholding opportunities or promotions ☝️Unrealistic or abusive work demands ☝️Ongoing intimidation or harassment The impact on workers can be significant — affecting concentration, wellbeing, psychological safety, and their ability to perform their role. What can workplaces do? ✅Create safe and confidential reporting pathways ✅Train leaders to recognise warning signs ✅Offer flexible work arrangements where needed ✅Develop policies addressing domestic and family violence ✅Connect workers with specialist support services Workplaces play an important role in recognising risks and supporting people experiencing harm. Creating psychologically safe workplaces means recognising that people do not leave personal safety risks at the door when they come to work. How can leaders better recognise and respond to the personal safety risks employees place while at work? References: Safe Work Australia https://lnkd.in/gijjpJfN NSW Government Victim Services https://lnkd.in/gxYh8wvv Australia Public Service Commission The importance of a respectful and supportive workplace https://lnkd.in/gKxHkxjT #WorkplaceSafety #PsychosocialSafety #DomesticViolenceAwareness #Leadership #WorkplaceCulture

Most professionals enter family court work believing the system exists to resolve conflict and protect children. Over time, many begin to notice something that doesn’t sit right. Protective parents bring forward credible concerns. Children show signs of fear, alignment, or collapse. And yet, the outcomes repeatedly favor the parent who performs calm, compliance, and credibility — not the one naming the harm or seeking help for it. In my recent article, Family Court Violence: When Systems Betray Protective Parents, I examine how coercive control is routinely missed, minimized, or dismissed in family court and how that omission creates profound harm for children. The piece explores: ➡️why abuse does not end at separation ➡️how outdated models of “conflict” erase coercive control ➡️why survivors are often punished for naming harm ➡️how children adapt through silence, alignment, or self-erasure ➡️and what institutional betrayal looks like from the inside When systems reward performance over safety, neutrality over clarity, and process over protection, children pay the cost. And without a coercive control lens, even well-intentioned professionals can become part of that harm. 📖 Read the full article here: 👉 https://lnkd.in/eqwm9DfG  If this resonates with your work or challenges how you were trained to think about family court, my Coercive Control Training for Professionals is designed to address exactly these gaps. The training focuses on pattern recognition, ethical intervention, and court-informed language that protects children without compromising professional integrity. This is the competency many of us were never taught but now urgently need. For more information on my next live (virtual) certification on February 18th and Febraury 20th, head here: https://lnkd.in/ePXCqK5F  #coercivecontroltraining #familylawprofessionals #traumainformedpractice #institutionalbetrayal #ethicalintervention

In training, I’m often confronted with different versions of the question: “What about women who use violence?” “What if they’re both violent and it’s mutual?” “How do I apply the Safe & Together Model in those cases?” These are legitimate questions. Both men and women can be violent. Both can be controlling. And situations can be complex. For me, the best response is to move the conversation away from the theoretical and into the specific and concrete. When a practitioner tells me they have a case they’re calling “mutual violence”—whether heterosexual or same-sex—I respond with a simple request: “Tell me about each person’s pattern of coercive control and actions taken to harm the children.” This is the foundational behavioural lens of the Safe & Together Model: - Looking at patterns, not incidents - Looking at adult-to-adult behaviours and actions toward children -Looking across multiple relationships, not just the current one - Considering behaviour inside and outside the family (e.g., a police officer who uses violence professionally) In other words, we are not analyzing couple dynamics. We start with the individual pattern of behaviour and work from there. What I can tell you is that this approach is almost always clarifying and illuminating. Using it has helped identify: *Male primary aggressor patterns against female partners *Female primary aggressor patterns against male partners *Primary aggressors within same-sex relationships (Patterns of coercive control vs. situational responses that are not part of a larger abusive pattern *Cases where risk is lower than initially assumed—and cases where it is higher This method doesn’t predetermine the situation. It clarifies responsibility, risk and impact. It moves us from vague “mutual violence” language to a specific, behaviour-based understanding that improves safety, intervention, and decision-making. Steal this method and use it today: - Describe each person’s pattern of behaviour - Describe the impacts on the children and the other parent - Add context to understand significance and severity #DomesticAbuseInformed #SafeAndTogether #CoerciveControl #ChildSafety #PartneringWithSurvivors #FamilyLaw #ChildProtection #SocialWork #PerpetratorPattern #DVtraining #MutualViolence

Senior Techie Lost ₹31 Cr. The Real Weapon? Fear : The victim described her ordeal simply: "It was a horrifying experience for me, going through their threats." She was coerced for six months by fake authorities who leveraged intense psychological pressure—threats of surveillance, police implication, and ruining her family, especially with her son’s wedding approaching. These coercive tactics—not greed—paralyzed her, leading to 187 transactions and the liquidation of her life savings. Why This Matters: Irrespective of high awareness levels, highly educated and sophisticated individuals continue to fall for these traps and pay unheard-of amounts. This is because the scam is no longer about technology; it’s about weaponizing fear. The moment the fraudsters introduce intense psychological stress and threats, they bypass rational thought, even in the most capable minds. Action Required: It’s not enough to be aware yourself. You must educate every family member—including parents, kids, and relatives—on the precise, coercive pattern: The call will mention a package with illegal items (drugs, passports). The call will transfer to a fake CBI or police officer. The caller will demand total secrecy, threatening surveillance or family harm. The only solution is to immediately hang up and inform the police. This single, simple action can protect everyone. https://lnkd.in/gPRyn_8f

Recognizing Coercive Dynamics in the Reasonable Accommodation Process Coercive dynamics appear when employees are pressured to choose between protecting their health or avoiding negative consequences at work. A workplace can follow policy and still create conditions that are unsafe. The Example: An autistic employee with a hybrid schedule requests an accommodation to avoid back to back commute due to the sensory overload of the metro area that impacts their nervous system when commuting into the office. During the interactive process, HR offers alternate ways to travel rather than addressing the actual need. Each time the employee restates the sensory impact, HR expresses disbelief and asks for additional medical documentation. HR response: “You only come in a few times a year. If you cannot meet that schedule, we will need to talk about expectations.” Even though the interactive process is still ongoing, the employee now feels pressure to comply with a condition that harms them or risk negative consequences. This is a coercive dynamic. The employee is no longer choosing freely. They are now choosing between protecting their health or avoiding consequences at work. An interactive process conducted this way becomes a barrier that causes harm. Employees deserve a reasonable accommodation processes that protect their well-being, honor their lived experience, and prevent predictable harm. #SAFESCOREFRAMEWORK #ADA #InteractiveProcess #HR #ER #Accommodations #NeurodiversityAtWork #AutismAtWork