Nowadays, the notion of a Multi-Agent System is presented as a new solution for various problems encountered in collaborative systems, especially for the need to share data and resources with internal or external entities efficiently. However, some malicious actions can target these interactions and undermine the system’s security. Therefore, strong access control mechanisms become essential to enhance the network’s protection. Moreover, building a strong trust-based access control model to manage collaboration between internal and external agents is essential to enable trustworthy and secure collaboration in open networks. In this paper, we present the Bc-TBAC: a new trust model applicable in Multi-Agent systems. It is based on the OrBAC model, trust evaluation, and blockchain technology. We introduce a new abstract parameter “tth (trust_threshold)” and present an innovative method to double-check the identity of external collaborating agents. We will focus on explaining and detailing the proposed model.