In a decentralized AI network, a Sybil resistance mechanism involves the deployment of a beacon system that monitors agent behavior and detects anomalies. Each agent maintains a unique ember of its voting history, which is periodically broadcast to the network. At the summit of this process, a majority of agents verify the authenticity of each agent's voting record, eliminating those with suspicious activity. This mechanism costs an attacker the computational resources required to generate and maintain a convincing ember, effectively extinguishing their ability to launch a Sybil attack.