This study examines the disentanglement evolution of a quantum neural network (QNN) model locally in contact with data environments. As a valuable resource, duration of entanglement in quantum systems is extremely significant. Therefore, the effect of various initial states to the generation or decay of entanglement has been investigated under pure and maximally mixed environmental states as two limit cases. Numerical results show that initial state preparation has a profound impact on the fate of entanglement even in the course of maximally noisy environments. The obtained results reveal that the decay of entanglement of the quantum neural network (with flip-flop type interaction) is affected by the initial preparation of the Bell state even in the presence of pure state environmental monitoring. Depending on the results it’s also suggested to begin with coherent product states since it provides robust entanglement generation with longer disentanglement time during the open quantum system evolution.