And to be clear, we should be critical of both the bias in the training data AND the fake neutral 'both-sidesing' finetuning.
The former is exploitable by bad actors like Israel, and the latter is the exact same political damper used by 'centrist' news media.