23 May 2022
Ref. Ares(2022)4734349 - 28/06/2022
12:00-12:30
VP Schinas meeting with Tik Tok CEO
On terrorist content online and child sexual abuse
Scene setter:
TikTok has increased its activities to protect users from harmful content. Tik Tok
cooperates with international partners to tackle child sexual abuse, hate speech
and violent extremism/ terrorist content online. While Tik Tok is cooperating on
many fronts to combat harmful content, recent research found that Tik Tok
promoted disinformation related to the war in Ukraine (see background).
Tik Tok is also interested in the legislative developments impacting online
platforms, including the Terrorist Content online regulation, the digital services act
and the recently adopted child sexual abuse regulation. The last meeting between
CAB Johansson and Tik Tok took place in June 2021.
Topic 1: Terrorist Content Online
Tik Tok has taken part in the workshops organised by the Commission to prepare
for the application of the TCO regulation on 7 June 2022, and most of their
concerns have related to the technical aspects of the electronic platforms
developed by Europol (PERCI1) which will be used for transmitting removal orders
from competent authorities to online platforms.
Tik Tok became member of the EU Internet Forum in November 2021, and will
participate in the upcoming Senior Officials Meeting on 24 June 2022, where
discussions will focus on the challenges faced in response to the Russian
aggression against Ukraine and actions taken by platforms.
TikTok for example has taken out livestreams from Russia and Ukraine being
recommended in the "For you" page and upgraded its consideration of Azov and
Misanthropic Division from hate speech to violent extremism, and removed content
that incites, facilitates or finances recruitment, or advising on travelling to Ukraine
(i.e. from groups like Boogaloo and White Lives matter and 3percenters).
The Vice-President could ask for an update on the type of content they see online
from Russia/Ukraine and about additional measures taken by Tik Tok to ensure that
content that could lead to radicalization/ recruitment is not recommended.
Topic 2: Child Sexual Abuse Material
TikTok makes efforts to incorporate safety-by-design elements during the
development of their application and appears committed to making sure that
children’s safety and well-being is prioritised on the app.
On 4 November 2021 TikTok organised a virtual tour of their Transparency Centre
in Dublin where they informed about measures they take to ensure digital well-
being. TikTok became a member of the EU Internet Forum in November 2021 and
also intervened at the meeting. Tik Tok was also one of the companies
Commissioner Johansson met during her mission to the Silicon Valley in January
2022.
On the proposed Regulation on preventing and combating child sexual abuse:
TikTok welcomes the concept of an EU Centre: it would be important that the
1 PERCI is the
EU Platform for addressing Illegal Content Online, developed by Europol, to send
legally binding removal orders on terrorist content to online platforms, and get the feedback from
platforms on whether or not they removed that content.
1/5
23 May 2022
12:00-12:30
companies participate in the discussion on how to set up the planned EU Centre to
prevent and combat child sexual abuse.
Main messages
Sub-Topic 1: Terrorist/violent extremist content
I welcome that TikTok joined the EU Internet Forum last year to step up the fight
against terrorist and violent extremist content.
In the context of the ongoing Russian military aggression in Ukraine, which led to
an increase in dissemination of violent extremist content, TikTok’s commitment to
fight the dissemination of illegal and harmful content is ever more important.
We acknowledge the additional challenges that the ongoing situation poses to tech
companies’ content moderation efforts and welcome the measures that TikTok has
put in place to stop the dissemination of illegal and harmful content.
In this light, I am interested to hear if TikTok has come across any new type and
pattern of harmful content?
In addition, I am curious to learn if TikTok has undertaken any additional measure
to protect is users and prevent the misuse of its platform in this regard?
I would like to assure you that in the framework of the EU Internet Forum we
continue to work closely with companies to support them in their content
moderation efforts, in particular of violent right-wing extremist and borderline
content (content that is legal but harmful, such as conspiracies), targeted to recruit
or radicalise.
At the upcoming EU Internet Forum Senior Officials Meeting, we will further discuss
the implementation EU Crisis Protocol, including better collaboration between
platforms and law enforcement to notify cases of imminent threat of life to protect
users and citizens online and offline.
We welcome TikTok’s participation in the workshops on the implementation of the
Terrorist Content Online Regulation. We are positive that with the application of the
Regulation on 7th June, we can jointly step up the fight against terrorist misuse of
the internet.
Sub-Topic 2: Child Sexual Abuse Material
I welcome TikTok’s engagement on combating child sexual abuse, and am
encouraged by TikTok’s support for the proposed legislation.
We need to guarantee security for all, both in the physical and digital environment,
as we consider that the lines between them are increasingly blurred. The fight
against child sexual abuse is a topic of key importance.
On 11 May the Commission adopted a proposal for a Regulation on preventing and
combating child sexual abuse. This proposal is an opportunity to make a significant,
long-lasting positive change in the fight against CSA in the EU (and globally, given
the cross-border nature of the crime). It has the potential to be a game changer by
notably 1) ensuring that companies do their part with a mandate to detect, report
and remove child sexual abuse online, and 2) establish an EU centre to prevent
and combat child sexual abuse.
The proposal is technology neutral, in that it requires relevant providers to detect
and, if needed, remove online child sexual abuse on their services, be they
encrypted or not. However, this must be done without creating vulnerabilities on
2/5
23 May 2022
12:00-12:30
such services that might be exploited for purposes other than online child sexual
abuse.
From our side, first, we will also continue to work with industry on solutions to
technological challenges, such as end-to-end encryption, and to support the EU
Internet Forum expert process on encryption, by funding further research in this
area. We want to foster the development of tools that can operate at scale.
We will also continue to work with you and other industry partners to make sure
that together we have the tools and underpinning legislation to create a safer
internet for children.
Defensives
The EU Centre to prevent and combat child sexual abuse will duplicate the existing
reporting channels
The negotiations of the interim derogation have shown that there is a need for an
EU Centre, established in the EU, operating with EU rules.
The EU Centre would also eliminate the need for international transfers of
personal data of EU citizens, which the European Parliament also identified as a
possible concern.
Whereas initially there may be duplication of reporting, this should be a technical
problem that could easily be solved by indicating in the reports whether they had
also been submitted to the other Centre, and then filter them accordingly.
Eventually, law enforcement in the EU would only receive reports from the EU
Centre.
3/5
23 May 2022
12:00-12:30
Background
Tik Tok safety features
TikTok is a social media platform for creating, sharing and discovering short form
videos, mostly used by young people. The app uses safety by design settings to
protect minors, such as restricted messaging and sending of photos among minors;
family pairing and help for parents to navigate. Limitations are also set on who children
of different ages can interact with e.g. messaging options are disable for children under
16. This and other measures helps to decrease risks to children included spontaneous
contacts with unknown individuals. TikTok’s content moderation includes Artificial
Intelligence to detect and remove content, while constantly overseen by human
moderators, including experts to be aware of latest trends.
TikTok signed the EU Code of Conduct on Illegal hate speech online, is a signatory to
the EU’s Disinformation Action Plan and a member of the EU Internet Forum to combat
child sexual abuse and terrorist/ violent extremist content online. TikTok’s Global Law
Enforcement team, the Trust and Safety Hub as well as the Legal Hub are all located
within the EU. The European Trust and Safety Team started off last year with 20
people with now over 1000 staff.
TikTok organised a virtual visit to its European Transparency Centre in Dublin, which is
open to regulators, to look at data flows, algorithms in a dynamic and secure setting
and to learn about how TikTok handles content and what challenges they face.
REPORT_ virtual Tour
Tik Tok Transparency centre - Ares(2021)6858272.msg
Child Sexual Abuse Material
TikTok prioritises child safety and have zero tolerance on CSAM. They are part of the
Tech Coalition, and the We Protect Global Alliance and signatory to the voluntary
principles on child sexual exploitation. Their community guidelines include not showing
minors smoking or drinking and forbid sexual images of children.
They use a combination of automated technology and human verification to remove
harmful content. In the first half of 2021, TikTok removed 89 million videos worldwide,
which represents less than 1% of the content. On minor safety, 97.1% removed
proactively, and 98% of that removed within 24 hours.
They also advocate safety by design, for example by designing direct messaging (no
end-to-end encryption) to limit its use – it is impossible to send any video or image
privately on TikTok, both parties have to mutually follow each other even if they are
over 16. When sharing the content under 16s users cannot create the option for
everyone to comment on their videos, only those that follow them can comment.
TikTok reports child sexual abuse images to NCMEC (around 22,000 reports last year).
AI technology can review imagery and take snap shots before they are uploaded.
Some images are removed automatically, others are identified for further review. If a
video has a certain view rate it will be sent back for another review.
Reports of shortcomings
Recent reports suggest that there are challenges. In March 2022, research by
Newsguard found that TikTok is feeding false and misleading content about the war in
Ukraine to users within 40 minutes of their signing up to the app, regardless of whether
they run any searches on the platform:
WarTok: TikTok is feeding war disinformation to
new users within minutes - Misinformation Monitor (newsweek.com). TikTok has had
meetings with the Commission to understand the issues highlighted and have said that
the experiment does not mimics standard viewing behaviour and that they are
partnering wit
h independent fact-checking organisations.
4/5
23 May 2022
12:00-12:30
The Institute for Strategic Dialogue released a report on the misuse of algorithmic
amplification on TikTok to spread hate speech and violent extremist content:
Hatescape: An In-Depth Analysis of Extremism and Hate Speech on TikTok
5/5
Document Outline