Ethical decision-making has become a prominent step in evolving digital decisions. It is often crucial to take mindful steps to maintain transparency in decisions, particularly when undertaking judicial practices in special missions using tiny edge-enabled devices. In fact, automated decision support systems often fail owing to biased data or complex AI solutions without sufficient edge-assisted systems in traditional approaches. In this article, a Blockchain-based Human-in-the-loop Facial Recognition (BHFR) is proposed that involves tiny microcontroller devices to process learning models and edge-enabled devices to execute permissioned blockchains that validate and perform ethical decisions, especially when culprits need to be penalized. BHFR aims at targeting special missions in identifying culprits in a collaborative manner. Experiments were carried out at the IoT Cloud Research laboratory to study the impact of involving tiny microcontroller devices nearer to sensor nodes and the efficiency of involving blockchains that combine multi-stakeholders’ confirmations. Additionally, the face enrollments and the corresponding training accuracy with memory consumption and predictions using tiny microcontrollers are reported in the article.