It is no wonder technology companies across every field are rushing to deploy mechanized content moderation. It has also remained silent when asked whether it would permit external review of its algorithms and their filtering decisions. We have absolutely no idea how well Facebook’s algorithms work or how much innocent content they are leading to be wrongfully removed. While noting that 38% of removed abusive content was flagged by its algorithms, Twitter notably did not provide any indication of its false positive rate. It is almost guaranteed that Twitter’s nascent moderation algorithms are subject to any number of implicit biases in their results.

Source: Forbes April 24, 2019 03:45 UTC