Size‐based indicators are used worldwide in research that supports the management of commercially exploited wild fish populations, because of their responsiveness to fishing pressure. Observational and experimental data, however, have highlighted the deeply rooted links between fish size and environmental conditions that can drive additional, interannual changes in these indicators. Here, we have used biogeochemical and mechanistic niche modelling of commercially exploited demersal fish species to project time series to the end of the 21st century for one such indicator, the large fish indicator (LFI), under global CO₂ emissions scenarios. Our modelling results, validated against survey data, suggest that the LFI's previously proposed policy target may be unachievable under future climate change. In turn, our results help to identify what may be achievable policy targets for demersal fish communities experiencing climate change. While fisheries modelling has grown as a science, climate change modelling is seldom used specifically to address policy aims. Studies such as this one can, however, enable a more sustainable exploitation of marine food resources under changes unmanageable by fisheries control. Indeed, such studies can be used to aid resilient policy target setting by taking into account climate‐driven effects on fish community size‐structure.