Neuroinflammation and the underlying dysregulated immune responses of microglia actively contribute to the progression and, likely, the initiation of Alzheimer's disease (AD). Fine-tuned therapeutic modulation of immune dysfunction to ameliorate disease cannot be achieved without the characterization of diverse microglial states that initiate unique, and sometimes contradictory, immune responses that evolve over time in chronic inflammatory environments. Because of the functional differences between human and murine microglia, untangling distinct, disease-relevant reactive states and their corresponding effects on pathology or neuronal health may not be possible without the use of human cells. In order to profile shifting microglial states in early AD and identify microglia-specific drivers of disease, we differentiated human induced pluripotent stem cells (iPSCs) carrying a familial AD PSEN2 mutation or its isogenic control into cerebral organoids and quantified the changes in cytokine concentrations over time with Luminex XMAP technology. We used partial least squares (PLS) modeling to build cytokine signatures predictive of disease and age to identify key differential patterns of cytokine expression that inform the overall organoid immune milieu and quantified the corresponding changes in protein pathology. AD organoids exhibited an overall reduction in cytokine secretion after an initial amplified immune response. We demonstrate that reduced synapse density observed in the AD organoids is prevented with microglial depletion. Crucially, these differential effects of dysregulated immune signaling occurred without the accumulation of pathological proteins. In this study, we used microglia-containing AD organoids to quantitatively characterize an evolving immune milieu, made up of a diverse of collection of activation patterns and immune responses, to identify how a dynamic, overall neuroinflammatory state negatively impacts neuronal health and the cell-specific contribution of microglia.