Recognizing members of the same social group as in-group individuals occurs swiftly and without conscious effort. The exploration of the neural processes underlying this process can be advanced through behavioral experiments designed to trigger naturalistic social interactions. Here we show a novel experimental approach to studying the formation of social bonds between two groups of mice: genetically identical, however unfamiliar with each other. Animals were placed in Eco-HAB, an automated environment that replicates natural conditions, for 21 days. The habitat was initially split into two, providing each group its own space for 7 days. Then groups were merged and allowed to interact freely. Our findings indicate that directly after combining the groups, the mice showed a preference for their original group over the unfamiliar one. Post-merger mice voluntarily spend more time with their original group members, though the extent of that behavior varied by individual. This changed, however, in the hours that followed; the mice began to follow out-group individuals more often. Next, the social dynamics began to alter, with some mice maintaining their initial social inclinations and others developing new affinities towards previously unknown individuals. Moreover, we also discovered that chemogenetic activation of PV cells in the prelimbic cortex decreased time voluntarily spent together with both familiar and unfamiliar conspecifics. To conclude, our research provides insights into how two distinct groups can come together to form a unified social unit. These findings lay the groundwork for further investigation into the neural underpinnings of forming new social connections.