Not only conversatives. Liberals are also upset about vaccines misinformation, election fraud, etc.
My point is that you have to 'train' the AI models somehow. FB introduces its own 'biases' on those models. They admitted this 'mistake' only because it affected a Canadian MP, otherwise I don't think they would have this for you or I. Their 'discrimation' policies are too broad for an automated system.
FB exhorts too much power over information.