The response of tropical cyclone activity to anthropogenic radiative forcing remains uncertain, with even the direction of the change uncertain in some respects (e.g., TC frequency), both globally and regionally. One important source of uncertainties is the Pacific zonal SST gradient. At the interannual time scale, this SST gradient, through the El Niño Southern Oscillation, is known to strongly influence global TC activity. Global climate models in CMIP5/6 generations project this SST gradient to weaken and lead to a more El Niño-like mean state in the future. Observations over the past several decades, however, show a strengthening of the SST gradient and thus a more La Niña-like mean state. While the observed strengthening of the SST gradient may be due to natural variability or merely an observational issue, some recent studies have marshalled evidence, backed up with modeling, to argue that the projected weakening is erroneous and a consequence of a common climatological cold tongue bias that has persisted in a few generations of global climate models. If the above argument is correct, at the transient forced response of Pacific SST over the upcoming decades will be towards a La Niña-like mean state, in contrast to the climate models. This means that the projected trends in TC activity from current state-of-the-art global climate models may be incorrect in some basins. In this presentation, we will report an initial investigation of the above problem using synthetic TCs from the Columbia tropical cyclone HAZard model (CHAZ) downscaled from CMIP6 models. Although all show El Niño-like forced responses, we will group the CMIP6 models/members based on the magnitudes of their climatological cold tongue biases, their historical trends of the zonal and meridional SST gradients, and the correlation between their trends and the observed one. For each stratified group, we will then evaluate SST gradient projections and how these projections affect the large-scale atmospheric and oceanic environment conditions that are important to TC activity and thus influence the forced trends in the CHAZ-CMIP6 downscaled TCs. This work will inform on how much a potential model bias towards the wrong sign of the tropical Pacific zonal SST gradient change matters for projections of global TCs.