This is a bit of a click-bait title, and is somewhat true -- but not in the way you think.
When FB was developing hashtags and content features, there was a concern that it is tough to find a middle-ground between 100% purely algorithmic generated content (i.e. leverage tastes data and supply articles suited to your interests) and 100% curated content (i.e. leverage tastes data, but only show articles from "reputable" sources and/or highly rated content, over shitty blog posts from amateurs.) The latter tends to do really well for engagement, as users can trust that the content they're seeing is reputable and popular. However the big concern that a number of us were raising was that tech companies have employees that are very biased in ways that they cannot control: they're younger, more liberal, and somewhat higher income, than the average FB user.
In my time there, I never heard of _explicit_ suppression of any viewpoints - with the exception of recent disagreements around the "Black Lives Matter" protests, the first time that I encountered a situation where disagreeing meant you were labelled a "racist" - but I can see why well-intentioned product and policy decisions led FB content down this hole.