PADAI Information

PADAI (Psychological Abuse Detection AI) exists to make hidden psychological abuse harder to ignore. Psychological abuse often hides in everyday messages. A phrase that looks caring or casual can carry manipulation, gaslighting, coercive control, guilt, humiliation, isolation, or emotional degradation. Seen alone, these messages may be dismissed as conflict. Seen in context, they can reveal harmful communication patterns that erode autonomy, safety, and well-being. PADAI is a survivor-informed AI tool and research platform for psychological abuse detection in text. It uses large language models, anonymized lived-experience data, prompt strategies, and responsible AI methods to detect, contextualize, and explain signals of hidden psychological abuse in written communication. The focus is nuance: pattern, sequence, repetition, and the difference between isolated words and a climate of domination. PADAI supports human judgment; it does not replace it. It can help survivors find validation and clearer language, assist lawyers and advocates with abuse evidence screening, support clinicians, NGOs, and counselors with context-sensitive analysis, and give researchers space to study AI for psychological abuse, forensic text analysis, gender-based violence prevention, and responsible model evaluation. Psychological violence is often recognized in law yet difficult for institutions to detect, prove, and repair. PADAI responds by combining lived experience, legal awareness, psychological insight, and advanced AI research. We are building tools, datasets, evaluation methods, and ethical safeguards to make invisible harm more visible without overstating what technology can decide. We welcome collaboration from survivors' advocates, lawyers, psychologists, clinicians, NGOs, researchers, journalists, policy makers, and funders advancing responsible AI for emotional abuse detection, coercive control detection, and safer understanding of abuse in text.