Decoding Negotiation Dynamics: The Impact of Opponent Identity and Privacy on Strategy, Deception, and Emotional Transparency in Human-Agent Interaction

Negotiation is a fundamental aspect of human-human and humanagent interactions, shaping decision-making and conflict resolution. As AI systems become increasingly embedded in these contexts, understanding how opponent framing (human vs. AI) and privacy decisions (webcam sharing) influence negotiation strategies merits investigation. This study examines their effects on deception and emotional engagement using the IAGO platform [8], where participants negotiate with an opponent framed as either human or AI while deciding whether to share their webcam. Results demonstrate that participants who withheld webcam data exhibited increased deceptive behavior, which positively influenced negotiation performance, though deception only partially mediated this effect. Although opponent identity did not significantly affect deception or success, participants exhibited higher emotional engagement when negotiating with a human opponent. These results underscore the necessity for privacy-aware, adaptive AI agents that foster engagement and ethical decision-making while aligning with human negotiation strategies.