It is hypothesised that ecological restoration in grasslands can induce an alternative stable state shift in vegetation. The change in vegetation influences insect community assemblages and allows for greater functional redundancy in pollination and refuge for native insect species. 2. Insect community assemblages at eight coastal California grassland sites were evaluated. Half of these sites had undergone restoration through active revegetation of native grassland flora and half were non-restored. Insects were collected from Lupinus bicolor (Fabaceae) within 2 × 2-m2 plots in spring 2017. Lupinus bicolor is a common native species that is used in California restoration projects, and home and state landscaping projects. 3. Ordination demonstrated that insect community assemblages were different between restored and non-restored sites. These differences were seen in insect functional groups as well as taxa-specific differences and were found to be driven by environmental characteristics such as non-native forb cover. 4. Functional redundancy of herbivores decreased at restored sites, while pollinators became more redundant compared with non-restored sites. The assemblages of the common species found at restoration sites contained more native insects than those found at non-restored sites, including species such as Bombus vosnesenskii. 5. Local grassland restoration has the potential to induce an alternative stable state change and affect insect community assemblages. Additionally, it was found that grassland restoration can be a potential conservation tool to provide refugia for bumblebees (Bombus), but additional studies are required to fully understand its broader applicability.