Twitter’s latest transparency report is out, and for the first time it includes data on requests to remove content posted by journalists, and requests resulting in the takedown of accounts promoting terrorism.
Read more: Facebook and Twitter face €50m fines if they don’t tackle hate speech
The report is Twitter’s tenth since it first began publishing them in 2012. It covers stats on government and non-government information requests, legal requests to remove content, trademark and copyright notices, and a new section called government ‘terms of service’ requests, which relates to non-legal requests around terrorist content.
Twitter’s very first transparency report preceded Edward Snowden’s revelations that governments were spying on their own citizens, and attempting to infiltrate social networks to scrape data or setup official arrangements to the same end. Five years later, the atmosphere in which Twitter has released this latest report is very different.
Although government overreach is still a public concern, Twitter and its peers have been under increasing pressure from Europe to step up its efforts to tackle hate speech and terrorist content.
As such, one of the key takeaways Twitter is pushing is that fact that from July 1, 2016, until December 31, 2016 (the period this latest report covers), 376,890 accounts were suspended for violations related to promotion of terrorism. This is published under its Government Terms of Service section, and is a steep increase considering 636,248 accounts have been suspended in total between August 1, 2015 and December 31, 2016.
It’s hard to tell what this truly means, without putting it context of the total number of flagged accounts in the same period (Twitter does not publish these stats, partially due to the problem of duplicate reports conflating the figures - for instance, Twitter received a total of 7 per cent more government requests for account information in the latest period versus the previous one, but this affected 13 per cent fewer accounts).
However, a key takeaway has been published alongside the 376,890 figure that appears to counter its implications, and assuages concerns about the social network being a lapdog to government panic. While 376,890 accounts were suspended, there were only 716 government requests made, covering 5,929 accounts. 85 per cent of these were ‘actioned’ after an investigation. The disparity in figures is because technology is doing the brunt of the takedown work - 74 per cent of the accounts removed were discovered by Twitter’s spam-fighting tools, 2 per cent were from government requests and the rest from user reports. The tools were first launched in 2015 and adapted to seek out content promoting terrorism. It targets bot accounts setup in droves, all tweeting similar propaganda, often verbatim. This section of the report will be expanded in the future to address concerns about terrorist content on the social network.
In December. the European Commission warned Facebook, Microsoft, Twitter and YouTube that if they failed to do more to tackle hate speech on their services, laws may be written to force them to do so. This followed on from each site failing to live up to the 24-hour deadline on removing flagged content they committed to, as stipulated in a voluntary ‘Code of Conduct on Countering Illegal Hate Speech Online’ signed in May 2016. Germany has lived up to the words of that threat, and is currently drawing up laws that could see internet giants fined up to €50m for non-compliance.
In the past week, Facebook, Twitter and Google also faced a Home Affairs select committee over their failure to tackle hate speech online. Google-owned YouTube was specifically singled out for capitalising on terrorist content through advertisements appearing alongside harmful videos. The general conclusion from MPs was that it was inconceivable tech companies with the power, wealth and technological capabilities of Google, Facebook and Twitter, could be incapable of doing better. The company spokespeople, including Twitter's head of public policy for the UK, Nick Pickles, agreed they should do more and argued they are protecting freedom of expression.
Twitter’s report, and the new section on Government Terms of Service requests, seems a direct response to those criticisms, and the start of a seemingly long and slow process of improvement.
Elswhere, the legal removals report now includes a section on requests to remove content posted by journalists and media organisations. It is an effort to show that Twitter still stands for freedom of speech and is one of its primary focuses as a platform for news-sharing. Turkey led the pack here, with 88 per cent of requests originating from the country. Only 88 requests were made in total worldwide, and Twitter only removed content in a handful of cases: one tweet pertaining to a violation of “an individual's personal rights in response to a court order” in Germany; 15 tweets and 14 accounts in response to 77 requests in Turkey relating to graphic images from a terror attack and tweets that violated a National Security Council decision after the failed coup. On any occasion where Twitter complied, it published the original court order on Lumen, a third-party site that studies cease and desist letters relating to online content.
Announcing the report, Twitter's director of global legal policy Jeremy Kessel said: "A lot has changed over the past five years, both at Twitter and around the world. But the need for transparency into government and company actions has never been more important given the current climate of continued crackdowns on freedom of expression and limitations on citizens rights around the globe.
"While we’re excited to have reached our tenth report, there is still plenty of work to be done.”
This article was originally published by WIRED UK
