Twitter claims increased enforcement of hate speech and abuse policies in last half of 2019

Twitter has given its biannual transparency reports a new home with today’s launch of the Twitter Transparency Center, which the company says was designed to make the reporting more easily understood and accessible. The launch was timed alongside the belated release of Twitter’s latest transparency report covering the second half of 2019. The company attributed the delay to the COVID-19 health crisis and its work in getting the new Transparency Center up and running. The report touts Twitter’s increasing efforts in enforcing its policies, including a 95% increase in accounts actioned for violating its abuse policy, a 47% increase in account locks and suspensions and a 54% increase in accounts actioned for violating hateful conduct policies, among others.

The company claims its ability to “proactively” surface content violations for human review has helped it increase enforcement of its rules, along with more detailed policies, improved reporting tools and other factors.

As a result, this period saw the largest increase in the number of accounts actioned under Twitter’s abuse policies — a metric that could speak to better technology, as Twitter claims, but also perhaps hints at the devolving nature of online discourse.

Meanwhile, Twitter attributed the increase in actions taken on accounts demonstrating hateful conduct, in part, to its new “dehumanization policy” announced on July 9, 2019.

Twitter increased enforcement of its rules in other areas during this reporting period, including the posting of sensitive media/adult content (enforcement actions were up 39%), suicide & self-harm (enforcement actions up 29%), doxxing (enforcement actions up 41%) and non-consensual nudity (enforcement actions up 109%). The only area to see a decline was violent threats, which saw a 5% decrease in the number of accounts actioned for policy violations.

Twitter also actioned 60,807 accounts for violating policies around regulated goods or services.

Online harassment has been a significant challenge for Twitter as it has grown. The social network now encompasses a wider swath of the general public, compared with its early days when tech enthusiasts knew it as twttr, a sort of public-facing SMS. Today, Twitter’s idealistic goal of being an “online public town square” is bumping up against the limitations of that model, which is also increasingly criticized as a flawed or even delusional sort of analogy for what Twitter has become.

Twitter, like much of social media, can over-amplify fringe beliefs, controversy and toxic content, to the detriment of conversation health. It can help polarize users’ opinions. And it serves as a breeding ground for cancel culture.

The company itself, as of late, seems to be waking up to the problem of putting the world together in one room to debate ideas, and the ramifications of amplifying misinformation that results.

It suspended accounts from the fringe conspiracy movement, QAnon, in July. It has also flagged and screened Trump’s tweets and briefly froze his ability to share misinformation. On the product side, Twitter rolled out a tool that lets users hide replies that don’t add value to conversations and, just last week, publicly launched a feature that lets users only tweet with friends and followers, instead of with the general public.

Abuse policy enforcement isn’t the only big change that took place in the last half of 2019. Government requests for user data also increased, Twitter found.

Twitter says that the U.S. made up the highest percentage of legal requests for information during the reporting period, accounting for 26% of all global requests. Japan was second, comprising 22% of information requests. Overall, government requests grew 21% in the period July 1 to December 31, 2019, and the aggregate number of accounts specified in the requests grew 63%.

Both metrics were the largest Twitter has seen since it began transparency reporting in 2012, it noted.

Twitter also saw 27,538 legal demands to remove content specifying 98,595 accounts — again, the largest number to date; 86% of these demands came from Japan, Russia and Turkey.

“Our work to increase transparency efforts across the company is tireless and constant. We will continue to work on increasing awareness and understanding about how our policies work and our practices around content moderation, data disclosures and other critical areas,” the company blog post about the new center explained. “In addition, we will take every opportunity to highlight the actions of law enforcement, governments, and other organizations that impact Twitter and the people who use our service across the world,” it noted.

More metrics, including those focused on spam, terrorism, child exploitation and extremism, are available on the new Twitter Transparency Center.