Supplementary written evidence submitted by Twitter (OHC0035)
Thank you again for providing Twitter with the opportunity to give oral evidence at the Home Affairs Select Committee recently. I am writing to follow up on several points made during the Committee.
Within Appendix I, additional details about information we provide to law enforcement with due legal process has been included. As discussed and agreed with the Committee Clerk, and as noted in your correspondence, we would be grateful if they were not shared in the public record or online.
We regularly receive data access requests from law enforcement agencies around the world. As detailed in our Transparency Report, we received 881 information requests from the UK government in the last reporting period (July-December 2018), and have a 75% compliance rate. This is one of our highest compliance rates globally.
We have clear and publicly available guidelines for law enforcement in place. They can be viewed on our website here. We also have a Legal Online Submissions Site in place to support Law Enforcement Agencies (LEAs) making requests to Twitter. We have received positive feedback from LEAs that this is a straightforward and efficient process for them to use and we are proud of the investment our teams have made in training LEAs how to use the online submissions site properly.
We will continue to notify law enforcement in situations where we have a good faith belief that there is an imminent threat of death or serious injury, and Twitter data is relevant to mitigating or averting that threat. Any requirement to proactively refer additional information would need to be pursuant to a legal framework that is consistent with international human rights law, particularly the rights to privacy and freedom of expression, as well as our obligations under other national laws. It is also an approach that could significantly increase the amount of data the police have to deal with, without any clear evidential value. To put this into context, all data about every Twitter user who breaks our rules would be an enormous dataset. In the second half of 2018, we took action on more than 500,000 accounts for violating our rules on abuse, hateful conduct and violent threats alone; and we have suspended more than 1.6 million accounts for the promotion of terrorism since August 2015.
It is worth noting, however, that in the case of content containing or promoting Child Sexual Exploitation, it will be removed without further notice and reported to the National Center for Missing & Exploited Children (NCMEC) under the established US legal framework. NCMEC distribute reports to law enforcement partners all over the world (e.g. NCA/CEOP in the UK). We have regular meetings with NCA CEOP, as well as maintaining direct contact on an operational basis where required.
You requested data on the number of accounts we have taken action on compared to the number of reports we receive. Our most recent Twitter Transparency Report (TTR), which was released after I gave oral evidence (July-December 2018), shared data across six Twitter Rules policy categories. 11,000,257 unique accounts were reported for possible violations of those Twitter Rules. 6,388 of those accounts were reported by known government entities. Out of these six categories, we actioned 250,806 accounts under hateful conduct policies, 235,455 accounts under abuse policies, 56,577 under violent threat policies, 30,907 under sensitive media policies, 29,824 under CSE policies, and 8,994 under private information policies.
The unique accounts actioned do not necessarily fall within the unique accounts reported dataset - this is for a number of reasons. During our review process, we may consider whether reported content violates aspects of the Twitter Rules beyond what was initially reported. For example, content reported as a violation of our private information policy may also be a violation of our policies for hateful conduct. If the content is determined to violate any Twitter Rules, it is actioned accordingly. Indeed, many of the reports that did not fall into these core categories were actioned under other policies, such as impersonation, unlawful use, or evasion of suspension.
We may also determine that reported content does not violate the Rules at all. And we often receive large volumes of false reports that bad faith actors have submitted in a coordinated effort to try and game our systems. As shared at the Committee hearing, 38% of abusive content that is enforced is surfaced proactively to our teams for review. This approach reduces the burden on users to report. As a result, the Unique Accounts Reported per policy categories above do not necessarily fall within the Unique Accounts Actioned dataset below.
We use the term ‘action’ to refer to our range of enforcement actions, which include possible account suspension or timeouts. We can take action at the Tweet, Direct Message, and account levels, and sometimes employ a combination of these enforcement actions.
While originally focused on requests submitted to Twitter by government actors, like court orders for information or content removal, over the years we have been working to expand our Transparency Report to also include more detail about the actions we take when enforcing the Twitter Rules. As part of our longstanding commitment to transparency, we will continue to evolve and expand this report, and we look forward to sharing these updates in the future.
Third, your Committee asked us to review several Tweets:
Finally, we have taken action on 95 white supremacist, nationalist or separatist organisations under our Violent Extremist Group policy. A number have been banned from our platform since 2017 - including Britain First, the English Defence League and National Front. We have a number of ongoing assessments, and continue to review these designations in light of both online and offline developments. As outlined in the latest TTR, there has been a 19% decrease in terrorist content on the platform since the previous reporting period. Of those suspensions, 91% consisted of accounts flagged by internal, purpose-built technological tools. The trend we are observing year-on-year is a steady decrease in terrorist organisations attempting to use our service.
Thank you again for inviting Twitter to participate in this Inquiry both in multiple oral evidence sessions and in written correspondence; please do not hesitate to be in touch if you have any further questions. We will endeavour to keep the Committee fully updated on any relevant action we take as regards extremist or terrorist content on the platform.
Katy Minshall
Head of UK Government, Public Policy and Philanthropy