As the need for privacy self-management in the Internet of Things (IoT) ecosystem grows, Privacy Assistants (PAs) have emerged as a solution to assist users. However, many existing PAs rely on static approaches, assume perfect knowledge of user privacy preferences, and expect complete around-the-clock responsiveness from users to elicitation prompts. Furthermore, they overlook the behavior of IoT devices and the information potentially available from surrounding PAs. As such, we designed PADOME, an adaptive PA that models the user’s privacy preferences and the IoT device negotiation strategy. PADOME integrates user privacy preference elicitation to better understand their privacy utilities, along with IoT device preference modeling and surrounding PA elicitation, to reduce uncertainty about the IoT device negotiation strategy. Designed using the DUNE framework and evaluated in the GEPARD simulation environment, PADOME demonstrates improved negotiation outcomes and higher agreement success rates.