Twitter has updated its hate speech guidelines to cover tweets that make dehumanizing comments. These comments treat “others as less than people” based on age, disability or illness. The changes follow updates from corporate policies last July that Twitter would remove tweets that dehumanize religious groups.
Previously, Twitter had issued a comprehensive ban on dehumanizing language in 2018 to complement existing guidelines for hate speech that cover protected classes such as race and gender. Since then, these dehumanization guidelines have been updated to take into account certain cases that were not addressed in their original rule set due to user feedback.
Now Twitter says tweets like the one in the image below will be removed when they are reported:
The company claims that reported tweets that violate these new guidelines but were released earlier today will be removed, but will not result in account freezes.
Twitter introduced guidelines to ban the dehumanization of speech for the first time in September 2018. At that time, Twitter asked for feedback and later announced that it had received more than 8,000 responses in more than 30 countries in just two weeks. Much of the feedback was about guidelines that were too broad. As a result, Twitter has started to demand certain types of speeches against certain groups and against its rules, from religion to age, disability and illness.
In a tweet, the company states that more groups may be protected by this policy:
More is coming. We continue to research and work with a global working group of external experts to address additional protected categories.
Further information on working with external experts can be found at: https://t.co/FTTW6k0DBJ
– Twitter Safety (@TwitterSafety) March 5, 2020
Twitter also doesn't allow transgender people to be named after the name they used before the transition. This is also known as "deadnaming". This policy was introduced in late 2018. The company announced in October 2019 that its automated moderation tools now flag and remove more than half of all improper tweets before users report them.