Twitter suspended climate activists' accounts for weeks by mistake

Environmental activists in Uganda say their Twitter accounts were suspended after what they believe was a coordinated campaign by opponents in the government there to silence them. 22-year-old Nyombi Morris has been actively protesting for the preservation of the Bugoma Forest, where the government has been allowing logging for commercial purposes.

Opaque takedowns — Morris's account and that of another activist were down for weeks and only reinstated following a report by The Verge. Twitter said that the accounts were caught up in its spam filter, but didn't offer any explanation as to why they were flagged, like whether or not someone triggered it through reports. Morris believes government officials abused Twitter's moderation system to get their accounts taken down following a television appearance about deforestation in Uganda.

Twitter is essential for Ugandans — More than 63 percent of the forests in Uganda have been logged in the past 25 years. The situation highlights the importance of having access to a public platform like Twitter for disseminating information in certain regions of the world. Uganda ranks near the bottom of the World Press Freedom Index. Without access to his Twitter account, Morris was suddenly cut off from his network of activists around the world. “We are not supported by our media. It is only social media which can support us, but if even on social media we are silenced — where will we go?”

Social networks like Twitter and Facebook have had to contend with tricky moderation issues that computers alone can't manage due the nuance of human language, but there are seldom enough human moderators to review the huge quantity of content that gets flagged. Especially during the coronavirus pandemic, Twitter is triaging and leaning more on automated moderation.

In response to a statement from Twitter last week saying it would remove tweets wishing for President Trump's death, users came out of the woodwork saying they're regularly threatened and harassed and the company doesn't take action against death threats or threatening content aimed at regular users.

AI is far from perfect — According to the company, more than 50 percent of harmful tweets are flagged by its automated tools before a user reports them, but Twitter admits its automated systems are far from perfect. "We want to be clear: While we work to ensure our systems are consistent, they can sometimes lack the context that our teams bring," Twitter said in a March blog post. The company is reporting longer wait times for human support due to the pandemic sending some staff home.

Activists say that Twitter can shape the perception of their work by either legitimizing it or suspending accounts, thereby suggesting their activity is harmful in nature.

Facebook has previously come under fire for deleting pages associated with anti-fascist pages, which critics said drew a false equivalency between violent white supremacist groups and those opposing them.