FIA (Female Intelligence Agency) is a pioneering organization dedicated to educating women in the emerging field of social discernment—an area focusing on the subtle power dynamics and “sneaky” manipulative maneuvers that can arise in everyday interactions. With a mission to help women build confidence in identifying and navigating complex social scenarios, FIA aims to equip individuals with the knowledge and skills necessary to protect their well-being and maintain healthy boundaries. Through workshops, seminars, and dedicated research initiatives, FIA brings to light the root causes of manipulative language, enabling women to recognize warning signs, assert themselves, and cultivate lasting personal agency.

Developed as a practical extension of FIA’s extensive research, ChatBoy aims to become a fun, interactive, and engaging tool that not only identifies harmful conversational behaviors but also teaches users how to respond calmly and effectively. The system leverages advanced NLP (Natural Language Processing) to detect potential manipulation and provides immediate feedback by offering short, guided interventions grounded in neurofeedback techniques. Central to our process is the commitment to rapidly build prototypes and subject them to rigorous, real-world testing. This iterative approach ensures that user feedback directly shapes the tool’s development, continually refining ChatBoy’s usability and accuracy.

Please check out our ChatBoy deck: https://docs.google.com/presentation/d/19BSt0OVwNAWV2k6gmnjM8INF7xTbvUowHR7FhVB6CwQ/edit#slide=id.p

Through user-centered design and agile development, ChatBoy is intended to evolve quickly: each iteration builds on the lessons learned from previous tests, integrating fresh insights into more robust features. This cyclical methodology reduces time-to-value for users, aligns with FIA’s agile mindset, and ensures that ChatBoy remains the most effective resource for women seeking to identify and disarm manipulative or coercive social tactics.

Rapid Prototype(s):

• Multiple rounds of quickly developed MVPs (minimum viable products) that users can test immediately to gather feedback.
• Emphasis on user-interface design, prompt responsiveness, and clear identification of manipulative language.

Core Research Integration:

• A robust database of known manipulative tactics and subtle power moves curated by FIA.

AI-Driven Chat Engine:

• We are going to start with a pool of generated questions and answers to start testing the concept
• Later, if there is time, we could implement an NLP-based model capable of detecting manipulative cues in real-time conversation.
• Integration with FIA’s proprietary data to refine detection accuracy and empower dynamic user responses.

User Feedback & Iteration Loop:

• Systematic collection of user data and survey insights for continuous improvement.
• Frequent sprints to incorporate feedback, fix bugs, and add new features within tight development cycles.