We use global coupled atmosphere-ocean-biogeochemistry models from the Coupled Model Intercomparison Project (CMIP5), under the RCP8.5 scenario, to show that the global interannual variability of the sea surface pCO (calculated as 1σ) could increase by 62 ± 22 % by 2090. This amplification is a consequence of a larger background pCO and a lower buffering capacity that enhance the response of pCO to surface temperature (T) and dissolved inorganic carbon (DIC) changes. The amplification is counteracted by a decrease in the sea-surface DIC interannual variability, which will likely cause a strong reduction on the pCO’s variability in the equatorial Pacific. The potential changes in seawater carbonate chemistry are simulated with higher consistency than those in the DIC and T anomalies driven by ocean circulation and biology. The changes in sea-surface pCO interannual variability are reflected in the ocean-atmosphere flux of CO and need to be accounted for future carbon projections.