Two months ago, the internet tried to banish Nazis. No one knows if it worked - The Verge

It’s also too early to say whether companies have definitively changed their moderation policies, or if we’ve seen all the ways that hate speech can manifest online. Last month, ProPublica found that Facebook inadvertently let advertisers target categories that included “jew hater” or “how to burn jews,” automatically generated by an algorithm based on user interests. Facebook restricted the categories, but Google and Twitter turned out to have similar options, letting people run ads against specific hateful phrases or people “likely” to engage with racial slurs. Google and Twitter also later apologized and promised they’d banned the offending options.