Twitter expands hate speech rules covering race, ethnicity and nationality

Twitter Inc has expanded its policy barring hateful speech to include "language that dehumanises people on the basis of race, ethnicity and national origin."

The company had planned to add new categories to the policy over time after testing to ensure it can consistently enforce updated rules.

Twitter banned speech that dehumanises others based on religion or caste last year and updated the rule in March to add age, disability and disease to the list of protected categories.

Civil rights group Color of Change, part of a coalition of advocacy organisations that have been pushing tech companies to reduce hate speech online, called the changes "essential concessions" following years of outside pressure.

In a statement, Color Of Change Vice President Arisha Hatch criticized Twitter for failing to update the policy before November's presidential election, despite repeated warnings by the advocacy groups about violent and dehumanising speech.

Hatch also said Twitter has declined to provide transparency into how its content moderators are trained and the efficacy of its artificial intelligence in identifying content that violates the policy.