Twitter has today published its tenth transparency report and as always it contains some revealing details.
The Twitter Transparency Report has been published every six months since 2012 and the data including in the latest version is for the second half of 2016. As always it covers a variety of areas including government and non-government information requests, legal requests to remove content, and trademark and copyright notices.
But it is a new section which is making most of the headlines this time round as, for the first time, Twitter has included information on what it calls “government ‘terms of service’ requests”. These relate to non-legal requests from Governments regarding terrorism-related content.
Unsurprisingly, this is an area in which Twitter has been coming under a significant amount of pressure as more and more governments focus on combatting terror promotion online. This report reveals that between July and December 2016 Twitter suspended 376,890 accounts for terms of service breaches relating to promoting terrorism on the site.
Over the 18-month period from August 2015 to December 2016 the total number of suspended accounts was 636,248 meaning this shows a significant increase over the most recent 6-month period. In the first six months of 2016, just 235,000 accounts were suspended for promoting terrorism and a year earlier the number was just 125,000.
However, Twitter claims that this increase is due to improvements in their own self-policing rather than a sizable increase in Government demands. Indeed, the report suggests that there has only been a 0.015% increase in Government requests for terrorism-related account suspension.
There was a total of 716 such requests over the reporting period relating to 5,929 accounts. Twitter eventually actioned 85% of these requests.
The social media company has responded to a report from by George Washington University in 2015 which highlighted the platform as being a major tool for ISIS to disseminate its messages of hate and intolerance. They have implemented new internal filters which they claim are having the desired effect. These figures seem to back that up.
Beyond terrorism, the transparency report has also revealed data about Government legal removals report now also includes a section highlighting content which has been specifically posted by journalists and media outlets.
It makes for bad reading for journalists in Turkey with 88% of these requests coming from the country. Twitter only complied with a handful of these requests and on each occasion it was at pains to publish the original court order which required it to do so. This is in line with Twitter’s corporate aim to be seen as a bastion of free speech.
Figures related to information and account removal were also pretty damning against the Turkish regime. Over the reporting period, they made 493 requests for account information and 2,232 requests to remove accounts or content.
This represents an increase of 76% and 25% respectively and clearly illustrates the Erdogan regime’s efforts to crack down on social media in the wake of last year’s failed coup in the country, which we have reported on previously.
Twitter claims to have not complied with any information requests and only 19% of their removal requests led to “some content [being] withheld.”
The USA led the way on information requests with 2,304 requests (more than 50% of the total) relating to 5,618 accounts with some information being given in 82% of cases. There were 681 requests from the UK relating to 1,017 accounts, with 79% getting some information.
The UK Government also had three accounts and 16 tweets withheld, with Twitter noting that one request was related to a court injunction related to information about a certain celebrity.