The feature had rolled out to users last November and it allows users to hide replies to their tweets behind an extra click. The idea of the feature is to lessen the power of the trolls and their ability to disrupt conversations.
Using the new API organizations which have to deal with large amounts of tweets could automatically decide to hide replies based on criteria relevant to them – like tweets containing blacklisted words or tweets detected as spam or being toxic.
The feature was already available in preview to a small number of developers who are now releasing tools based on the API. Jigsaw already integrated it in their Perspective API allowing you to automatically hide tweets which are scored as being toxic, while reshuffle developed scripts which will allow you to hide tweets based on a list of keywords or a username.
The feature does have it’s controversies as it could be potentially used to silence critics or to shut down people trying to refute misinformation. Nevertheless adoption is growing. The API is currently available through Twitter Developer Labs and any approved developer can apply for it.