Introduction. The active integration of generative neural networks into the educational environment creates new hygienic challenges associated with risks to the speech development and psychophysiological health of schoolchildren. A comprehensive assessment of this impact is needed to develop evidence-based preventive measures. The purpose a comprehensive analysis of the factors influencing the impact of generative neural networks (text neural networks, TNNs) on the speech culture and psychophysiological health of students in grades 6-11. Materials and methods. A sociological study was conducted by surveying 185 students. An original questionnaire was used to assess the frequency, purposes, and strategies of TNN use, the attitude of the social environment, and self-assessment of changes in linguistic competencies. For statistical data processing, correlation analysis, cluster analysis (k-means), Pearson's χ² test, Mann-Whitney U test, and Fisher's exact test were applied using the R environment (v.4.3.1). Results. A high involvement of schoolchildren in using generative neural networks was established (88.2%). Four behavioral patterns of working with generated text were identified, forming a continuum from active analytical modeling (21.2%) to passive direct appropriation (5.9%). It was shown that the developmental effect and hygienic risks are mediated by the depth of cognitive processing of the material. Statistically significant relationships confirm that high frequency and covert use of generative neural networks correlate with the emergence of a "neuro-style", difficulties in initiating independent work, and decreased self-efficacy, which is considered a risk factor for psychophysiological health. Cluster analysis identified three user types, among which "Dependent Pragmatists" (29%) represent a target group for hygienic intervention. Conclusion. The impact of text neural networks on schoolchildren is ambiguous and determined by the interaction of internal cognitive strategies and external social regulation. To minimize risks (digital dependency, cognitive offloading, leveling of speech style) and maximize the potential of generative neural networks as a development tool, it is necessary to develop and implement a system of hygienic standardization and pedagogical support aimed at forming conscious, reflective practices of interaction with the technology.