Twitter’s AI tackles Terrorism

Twitter has launched its 11th biannual Twitter Transparency Report. In March of 2017, Twitter introduced a new section covering government TOS reports, but it was limited to government requests to remove content under their prohibition of the promotion of terrorism. Twitter have now expanded this section to cover a broader array of categories under the Twitter Rules, adding: abusive behavior, copyright, and trademark. Twitter have announced more on their transparency rules as shown below:

Removing Terrorist Content

A little more than 2% of global reports, fell within the promotion of terrorism category. Under this category of reports, they are referring to accounts that actively incite or promote violence associated with internationally recognized terrorist organizations, promote internationally recognized terrorist organizations, and accounts attempting to evade prior enforcement.

Twitter’s continued commitment to eliminate such activity has resulted in an 80% reduction in accounts reported by governments compared to the previous reporting period of July 1, 2016 through December 31, 2016. Notably, government requests accounted for less than 1% of account suspensions for the promotion of terrorism during the first half of this year. Instead, 95% of these account suspensions were the result of their own proprietary AI tools, up from 74% in our last Transparency Report. Between accounts surfaced by both their internal tools and government reports, Twitter removed 299,649 accounts during this reporting period, a 20% drop from the previous reporting period. Notably, 75% of these accounts were suspended before posting their first Tweet.

Twitter have suspended a total of 935,897 accounts for the promotion of terrorism in the period of August 1, 2015 through June 30, 2017. For more details, please refer to their latest government TOS reports section.

Increasing Transparency

Twitter also updated the remainder of their biannual report with data covering January through June, 2017. During this time, Twitter received 6% more global government requests for account information that affected 3% fewer accounts than in the previous period, including requests that originated from four new countries: Nepal, Paraguay, Panama, and Uruguay. In addition, they  received approximately 10% more global legal requests to remove content impacting roughly 12% more accounts compared to the previous reporting period. These included requests from nine new countries: Bahrain, China, Croatia, Finland, Nepal, Paraguay, Poland, Qatar, Ukraine, and Uruguay.

Broadening and Deepening Our U.S. Transparency Reporting

Twitter also announced they’ve continued to expand their U.S. country report. In this latest version, they’ve included a breakdown of California state information requests at the county level, with plans to introduce this section to other states in future reports. Notably, Los Angeles County came in as the top requesting county, submitting 44% of total California state information requests. Providing this increased level of granularity should further help their users get a better idea of how frequently their local authorities seek user account information.

U.S. Law and #Transparency

As in past reports, Twitter is not reporting national security requests in the bands mandated by the USAFA. Where gag orders on National Security Letters have been lifted, they have included the actual numbers of requests. With regard to other national security requests they may have received, they continue to litigate government restrictions preventing them from sharing more granular information in their case Twitter v. Sessions (f.k.a. Twitter v. Lynch). In February 2017, a hearing was held on the government’s motion for summary judgment, where the U.S. government argued that the case should be decided in its favor because it had met the applicable requirements for showing that Twitter’s 2014 draft Transparency Report was properly classified. On July 6, 2017, the court denied the government’s motion for a summary judgment finding that the government submission did not satisfy the requirements of the First Amendment. On September 5, 2017, the government filed a motion for reconsideration, motivated primarily by the 9th Circuit decision in  EFF’s NSL litigation. Twitter will be filing its opposition to the government motion shortly.

Twitter state that they will continue to act on their commitment to #transparency.


Related posts

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.