Mechanical weed-control methods using autonomous weeding machines provide an option for farmers to reduce the use of herbicides while efficiently dealing with weed issues. However, there is limited knowledge regarding the effects of advanced mechanical weed control technology on soil bacterial communities in paddy fields. In this study, a field experiment was conducted in southern China in 2021. High-throughput 16S rRNA gene sequencing was employed to sequence bacterial communities in soil samples under three treatments: mechanical weeding, chemical weeding, and no weeding. The objective was to investigate the impact of different weed-control methods on the diversity, composition, and co-occurrence patterns of soil bacteria. While mechanical weed-control method did not significantly alter bacterial alpha diversity, it specifically increased the relative abundance of the Bacteroidota phylum. Co-occurrence network analysis further showed that mechanical weed-control treatment enhanced network complexity and the proportion of positive interactions among bacteria, likely due to the physical intertillage effect of the weeding machinery. Our one-year study indicates that mechanical weed control can promote beneficial shifts in soil bacterial interactions without reducing diversity, offering a novel microcosmic perspective for selecting appropriate weed-control methods in agricultural management and sustainability of soil ecosystems.