In this paper, we understand the shortcomings of the current mainstream IoT privacy protection methods through analysis, and in this way, we propose an evolutionary and signaling game model for IoT privacy protection. The model analyzes the stabilization trend of IoT platform penalty coefficients on privacy protection and provides protection strategies. Combining the implications of the signaling game model, the degree of IoT privacy protection is measured using the Bayesian equilibrium solving algorithm. Simulation experiments are conducted to evaluate the specific effect of the model on IoT privacy protection. The increase in the detection rate of the model accelerates the convergence of the probability of malicious nodes, e.g., when the detection rate increases from 0.7 to 0.9, the convergence time is reduced by about two stages. The larger the penalty amount of the IoT platform, the model recommends more aggressive protection strategies, and the probability increases from 0.16 to about 0.4. The game parameters of the model reflect the malicious behavior in IoT, and the trust level affects the game parameters. The model in this paper reduces the attack gain by 4% to 10% compared with the comparison model when the fixed defense gain is 1500, which can better reflect the influence of protection signals on the attacker’s actions.