[Docs] [txt|pdf] [Tracker] [Email] [Nits]

Versions: 00

INTERNET-DRAFT                                                 N. Elkins
Intended Status: Informational                                      EDCO
                                                                B. Shein
                                                   Software Tool and Die
                                                              V. Bertola

Expires: April 20, 2019                                 October 17, 2018

           Human Rights Considerations of Internet Filtering


   This document is a survey of the filtering of content.  The focus is
   on the human rights involved as cited in the Universal Declaration of
   Human Rights" which is one of the foundational documents for HRPC.
   The recent years have seen an increase in content filtering for a
   variety of reasons including to further the aims of governments who
   wish to maintain their rule and suppress dissent but also to enforce
   cultural norms, human rights and compliance with the law.  Filters
   also exist for security (botnets, malware etc.), user-defined
   policies (parental control, corporate blocking of social networks
   during work time, etc.), spam control, upload of copyrighted material
   and other reasons. This document is based on several real world
   considerations: the existence of national and regional sovereignty,
   Internet Service Providers (ISPs) and Content Distribution Networks
   (CDNs) that provide connectivity and content hosting services, Over-
   the-top (OTTs) and Content Delivery Platforms (CDPs) that play a
   disproportionate role in capturing the attention and "eyeballs" of
   many of the users of the Internet.

Status of this Memo

   This Internet-Draft is submitted to IETF in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

Elkins                   Expires April 20, 2019                 [Page 1]

INTERNET DRAFT                  ifilter                 October 17, 2018

   The list of current Internet-Drafts can be accessed at

   The list of Internet-Draft Shadow Directories can be accessed at

Copyright and License Notice

   Copyright (c) 2018 IETF Trust and the persons identified as the
   document authors. All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document. Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document. Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Elkins                   Expires April 20, 2019                 [Page 2]

INTERNET DRAFT                  ifilter                 October 17, 2018

Table of Contents

   1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . .  4
   2 Content Filtering by States and Public Authorities . . . . . . .  5
     2.1 Filtering to Prevent Freedom of Assembly or Information  . .  6
     2.2 Filtering to Enforce Cultural Norms  . . . . . . . . . . . .  6
     2.3 Filtering to Prevent Violence  . . . . . . . . . . . . . . .  7
     2.4 Child Pornography  . . . . . . . . . . . . . . . . . . . . .  7
     2.5 Unauthorized Gambling and Illegal E-Commerce . . . . . . . .  7
     2.6 User Generated Content (UGC) . . . . . . . . . . . . . . . .  8
   3 Content Filtering by Internet Service Providers  . . . . . . . .  8
     3.1 Filtering for Network and Computer Security  . . . . . . . .  9
     3.2 Filtering on Behalf of the User  . . . . . . . . . . . . . .  9
     3.3 Filtering for Commercial Reasons . . . . . . . . . . . . . . 10
   4 Content Filtering by Platforms Providing Content and Services  . 10
     4.1 Enforcing Cultural Norms . . . . . . . . . . . . . . . . . . 11
     4.2 Blocking Extremist Activity  . . . . . . . . . . . . . . . . 12
     4.3 Blocking Activity Inciting Violence  . . . . . . . . . . . . 13
     4.4 Copyright Protection . . . . . . . . . . . . . . . . . . . . 13
     4.5 Filtering for Network and Computer Security  . . . . . . . . 14
     4.6 Content Filtering by End-Users . . . . . . . . . . . . . . . 14
   5 Security Considerations  . . . . . . . . . . . . . . . . . . . . 14
   6  IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 14
   6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
     6.1 Normative References . . . . . . . . . . . . . . . . . . . . 15
     6.2 Informative References . . . . . . . . . . . . . . . . . . . 15
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 16

Elkins                   Expires April 20, 2019                 [Page 3]

INTERNET DRAFT                  ifilter                 October 17, 2018

1 Introduction

   This document explores the use cases and history of filtering of
   content at a protocol and category level, grouping them by type of
   entity.  The focus is on the human rights involved as cited in the
   Universal Declaration of Human Rights" [UDHR] which is one of the
   foundational documents for HRPC. However, any case of content
   blocking has an impact on online expression, thus the document tries
   to provide a complete picture of all the reasons and mechanisms that
   lead to the filtering and removal of content from the Internet.

   The recent years have seen an increase in content filtering by for a
   variety of reasons. States, through different legal instruments and
   public authorities, require the blocking of Internet content with
   different aims; undemocratic governments may wish to maintain their
   rule and suppress dissent, but also democratic governments use
   blocking to enforce cultural norms, human rights and compliance with
   the law .  Filters are also widely used by network operators and
   Internet access providers for security (stopping botnets, malware
   etc.), to implement user-defined policies (parental control,
   corporate blocking of social networks during work time, etc.), to
   reject spam and for other reasons.

   Over-the-top (OTTs) [WikiOTT] and Content Delivery Platforms (CDPs) -
   providers like Facebook, Google (YouTube), and Twitter that
   distribute streaming media or other content and services as a
   standalone product directly over the Internet, bypassing
   telecommunications and connectivity providers - implement filters to
   prevent the upload of copyrighted material or other content that
   infringes their policies; in some countries, such filters are
   mandated by law. End users also want to apply content filters or
   content classification schemes at the edge of the network, for
   example, to protect underage users of the local network or to prevent
   the risk of reaching dangerous and inappropriate websites by error.

   While filtering usually attracts the highest attention, there are
   other ways to discriminate content that could be employed, leading to
   similar results. For example, an access provider may isolate the
   traffic directed to a specific website or service and slow it down,
   or apply additional fees for it, up to the point where users desist
   trying to connect to that destination. Content tagging can also
   constitute a weaker content discrimination system; even if the
   content remains accessible, marking it as dangerous or unsafe with
   prominent advance warnings will discourage users from accessing it.

   Some call all content filtering "censorship".  For example, the
   Internet Draft [Censorship] defines blocking of content as:
   "Censorship is where an entity in a position of power - such as a

Elkins                   Expires April 20, 2019                 [Page 4]

INTERNET DRAFT                  ifilter                 October 17, 2018

   government, organization, or individual - suppresses communication
   that it considers objectionable, harmful, sensitive, politically
   incorrect or inconvenient.  (Although censors that engage in
   censorship must do so through legal, military, or other means, this
   document focuses largely on technical mechanisms used to achieve
   network censorship.)"

   We find the use of the word "censorship" in this context to have
   purely negative connotations.  That is, the implication of using the
   word "censorship" implies that filtering of content is always "bad"
   or as acting against important human rights.   The reality of the
   situation is that filtering of content is done for many reasons,
   including several which may be regarded as "good":  acting to
   preserve human rights either directly, when hateful and violent
   content is removed, or, in the case of security filters, indirectly,
   by providing safer Internet access that can encourage users to spend
   more time and energies online and enjoy all the deriving
   opportunities for education, free speech and assembly.

   All in all, a balancing of rights is often at stake, and the right to
   free expression of a content creator is not the only right that has
   to be considered and protected. We thus feel that the entire subject
   needs a more nuanced and careful examination, trying to establish
   principles, guidelines and technical protocols that can increase
   transparency and user control over these practices, allowing users to
   distinguish between the bad and the good uses of content
   classification and filtering schemes.

2 Content Filtering by States and Public Authorities

   States may filter content through several legal instruments and by
   order of different public authorities.

   In democratic countries, cases of content blocking are usually
   defined by a law and justified by an appropriate balancing of rights
   (see section 2.2) and needs/benefits. Depending on the country, the
   law may delegate to a specific public authority - either independent,
   or part of the government - the power to order blocking of websites
   and other content, or such power may be deferred to court orders
   following due legal process. In authoritarian countries, such legal
   basis and processes are often missing, and the blocking is more
   focused on protecting the authority of the ruler (see section 2.1).

   In technical terms, the filtering can either be applied at the IP
   address level, via firewall rules or routing alterations, or at the
   DNS level, by altering the results of the queries for the blocked
   names. The latter method is more precise, avoiding to block all other
   websites and services hosted on the same IP address, but is also

Elkins                   Expires April 20, 2019                 [Page 5]

INTERNET DRAFT                  ifilter                 October 17, 2018

   easier to circumvent for end users; thus, democratic countries
   usually prefer the latter method while undemocratic ones generally
   prefer the former.

   The procedure to apply the filters usually involves the appropriate
   public authority sending a list of the blocked IP addresses and/or
   DNS names to all the country's Internet access providers, requiring
   them to implement it on their routers and/or DNS resolvers. Providers
   not complying with these requests usually are subject to fines, to
   the cancellation of their license to operate (in countries where such
   license exists) or even to the penal prosecution of their legally
   responsible managers.

2.1 Filtering to Prevent Freedom of Assembly or Information

   What is sometimes informally called "censorship", has to do with the
   action of some governments to block websites that promote dissent and
   counter-information and organize protest actions and assemblies to
   contest the government, or even platforms such as Facebook or Twitter
   which might enable dissidents to organize protests.

   Other filtering is done to suppress knowledge of people who
   participated in protest movements being harassed, jailed or even
   killed.  Some governments actually shut down the Internet altogether
   to prevent any witnesses to unfortunate activities.

   These activities may all be regarded as acting against basic human
   rights in [UDHR].

2.2 Filtering to Enforce Cultural Norms

   Some filtering is done via legislation to enforce cultural norms,
   such as blocking sites which promote totalitarian and violent
   ideologies or falsify history and news in ways that attack and
   endanger certain parts of society.

   For example, in several countries the advocacy of totalitarian
   regimes such as nazi-fascism and communism, or of racist ideas and
   practices against religious or ethnic minorities (Holocaust denial,
   racism against people of African origin, etc.), is forbidden by law.
   While websites located inside the country can be physically taken
   down, the groups promoting these ideas often use anonymous hosting
   services in foreign countries, thus making blocking the access at the
   Internet provider level the only instrument available to these
   countries to enforce these laws.

Elkins                   Expires April 20, 2019                 [Page 6]

INTERNET DRAFT                  ifilter                 October 17, 2018

   In the general balancing of rights, this type of content - which may
   be seen as disinformation, and is generally used to promote
   undemocratic practices and discrimination against specific minorities
   and ethnic groups - is often considered extremely harmful to the
   safety and rights of the affected minorities and to democracy and
   public order in general, up to the point of overcoming the free
   speech rights of the content authors.

   This kind of rights balancing also depends on cultural norms, with
   countries such as the United States giving priority to the free
   speech rights even of hateful authors, and countries in Europe and
   Asia giving priority to the general safety and social peace. Thus,
   the related filtering practices have to be applied by country,
   depending on the nationality of the end user and on the applicability
   of jurisdiction.

2.3 Filtering to Prevent Violence

   As an extension of the previous case, filtering often also applies to
   content inciting violence and promoting terrorism, or making violence
   easier. Its objective is to protect the right to safety of the
   general population.

   The EU wishes to fine Facebook, Google, etc. for problematic content.
   [BBCTECH] reads in part:

   "If authorities flag content that incites and advocates extremism,
   the content must be removed from the web within an hour, the proposal
   from the EU's lead civil servant states. Net firms that fail to
   comply would face fines of up to 4% of their annual global turnover."

   In the United States, a federal court has issued a temporary
   injunction against publishing plans for 3-d printed guns on the

2.4 Child Pornography

   Another type of content which is often blocked is child pornography,
   as a way to discourage the exploitation of children for sexual
   reasons and protect their safety.

   [Child-Porn] A number of countries will obtain the IP addresses of
   visitors to child pornography sites.  They will attempt to tie these
   IP addresses to actual human beings so that they can be prosecuted.

2.5 Unauthorized Gambling and Illegal E-Commerce

Elkins                   Expires April 20, 2019                 [Page 7]

INTERNET DRAFT                  ifilter                 October 17, 2018

   In most countries, certain services are regulated and thus a license,
   often connected to the payment of specific taxes and fees, is
   required before being allowed to offer them online. While there may
   also be an economic motivation to this, such regulation is generally
   justified as protecting the safety and health of the population.

   Among the most commonly regulated businesses are:

   - Gambling
   - Weapons
   - Medical products and drugs requiring a doctor's prescription
   - Alcohol
   - Cigarettes and tobacco

   Some countries - for example Italy [ITALY-REG] - use content
   filtering to prevent access to websites offering these products for
   sale without meeting the country's regulation and/or without having
   paid the appropriate taxes and fees

2.6 User Generated Content (UGC)

   Legislation to attempt to ensure that User Generated Content (UGC)
   does not violate copyright laws has been proposed. [EUCOPY]

   The summary is:

   " Tech giants must pay for work of artists and journalists which they

   Small and micro platforms excluded from directive's scope

   Hyperlinks, "accompanied by "individual words" can be shared freely

   Journalists must get a share of any copyright-related remuneration
   obtained by their publishing house"

3 Content Filtering by Internet Service Providers

   Internet Service Providers (ISPs) provide access to the Internet to
   the general public. As such, they are usually required to apply any
   State-mandated filters, depending on the applicable jurisdiction, as
   described in section 2.

   However, there are additional cases in which ISPs implement
   filtering, or weaker content discrimination methods, on their own -
   they will be described in this section.

Elkins                   Expires April 20, 2019                 [Page 8]

INTERNET DRAFT                  ifilter                 October 17, 2018

3.1 Filtering for Network and Computer Security

   Most of the common threats to the security of the Internet, both in
   terms of network security and of security of the end users and of
   their devices, are based on connections to unsafe websites and
   services - either services that have been designed for malicious
   purposes since the beginning, or legitimate services that have been
   cracked and infected with malicious software.

   For example, phishing relies on leading the user's browser to a
   forgery of the website of one of the user's suppliers, like his bank
   or her utility provider. Malware, such as ransomware and viruses, is
   commonly spread by connecting the user's browser to an infected
   website that downloads the executable to his device and launches it.
   Botnets rely on stable connections between the clients on user
   devices (often millions of them) and one or more "command and
   control" hosts which move over time.

   To counter these attacks and protect their users and their network,
   ISPs often acquire timely lists of malicious hosts from specialized
   providers and make them inaccessible by filtering them at the
   connection level, either by IP address or by DNS name.

   This practice is becoming even more common and more useful as the so-
   called Internet of Things (IoT) gains adoption. IoT devices usually
   are strongly automated but have very little computing power, security
   features and update capabilities, making them very vulnerable to
   exploits and takeovers. Thus, protecting the home network rather than
   the individual device becomes the most viable solution for the
   security of the Internet.

3.2 Filtering on Behalf of the User

   In some cases, the end users actually desire that some content is
   filtered out and made inaccessible, so that they cannot reach it even
   by mistake. Three common cases are:

   -Security filters: The user explicitly asks the ISP to filter out
   malicious websites, as per the previous section.

   -Parental control filters: From [UK-Controls] The user asks the ISP
   to block content which is not deemed safe for children. This block is
   usually customizable by each user, depending on their own desires,
   and is requested by families with children accessing the Internet
   from their home network. In some countries, the provision of this
   service by the ISPs is either mandated by law or required by industry
   self-regulation efforts.

Elkins                   Expires April 20, 2019                 [Page 9]

INTERNET DRAFT                  ifilter                 October 17, 2018

   -Productivity filters: The user - typically a corporate network
   administrator - asks the ISP to block content which is inappropriate
   or disallowed on the workplace, as it would endanger the corporate
   network or reduce productivity. This content usually includes social
   networks, sports and leisure websites, etc.

   These filters can be provided for free, included in the Internet
   access service, or can constitute a specific additional service
   requiring opt-in and the payment of an additional fee. Services of
   this type are commonly available in several European countries, often
   with millions of customers.

3.3 Filtering for Commercial Reasons

   Some ISPs provide limited Internet access services that only allow
   access to specific types of applications (instant messaging, for
   example) or do not include access to specific types of applications
   (video streaming, for example). In these cases, connections to the
   disallowed content are blocked or slowed down significantly. This
   kind of filtering could also depend on specific partnerships - for
   example, an ISP may encourage its users to use a specific search
   engine by slowing down the connections to the other ones, in exchange
   for monetary compensation by the preferred search engine.

   Due to concerns over the market and competition impact of these
   practices, including potential limitations of user rights, they have
   been made illegal in some countries, upholding the so-called "network
   neutrality" principle.

4 Content Filtering by Platforms Providing Content and Services

   In addition to filters at the edge of the Internet, enforced by ISPs
   either on behalf of the State or on their own, those that manage the
   content and its delivery inside the network filter content as well.
   Again, this may happen because of their decisions, or because these
   companies are incorporated to do business under the laws of one or
   more nation-states, and therefore are subject to the regulations of
   such nation-states.

   This kind of filtering happens under several forms. For over-the-top
   and content delivery platforms (OTTs/CDPs), content may be examined
   and blocked, often automatically, when the user uploads it onto the
   platform, or may be verified and removed following a request by other
   users or after a court order. In some cases, for example in search
   engine results, the content will not be blocked, but will be marked
   as unsafe with a prominent warning discouraging the user to proceed
   with the connection, or will not be shown unless the user disables
   the default "safe" mode.

Elkins                   Expires April 20, 2019                [Page 10]

INTERNET DRAFT                  ifilter                 October 17, 2018

   Many of these platforms also employ policies that lead to the
   exclusion of a user from the platform after a certain number of
   breaches to acceptable content guidelines, thus silencing the user
   permanently (though users may try to open a new account, but losing
   all their existing followers and connections).

   Content Delivery Networks (CDNs) and hosting providers also have the
   option of taking websites down entirely by shutting down their web
   service (see section 4.4 for an example). Similarly, domain name
   registries and registrars may make content temporarily inaccessible
   by discontinuing the domain name registration for the hostname used
   in URLs, though, differently from CDNs and OTTs, they cannot actually
   remove the content from the Internet.

   While some of these filters depend on applicable laws, in most cases
   the content guidelines are self-imposed, and may err on the side of
   content restrictions to reduce the legal risk for the platform, at
   the cost of reducing the user's chances to speak. In some cases these
   filters are managed by algorithms and artificial intelligence
   applications, making it hard for the user to even understand why the
   content has been blocked; often, no explanation and appeal mechanism
   is provided, or the appeal is untimely and ineffective.

   Even when laws apply, given the global nature of these platforms, the
   applicable laws are often not those of the user's own country, and it
   is almost impossible for the user to exert any legal rights or
   request due judiciary process.

   Additionally, the more the specific service is globally consolidated
   in the hands of a few big competing players and the more these
   filters become impactful; particularly in the case of OTT social
   networks, the termination of an account often cannot be adequately
   replaced by a new account on any competing service or even on the
   same one.

   Some examples of similar situations follow.

4.1 Enforcing Cultural Norms

   The article from the Guardian [FBNorms] expresses the thoughts of the
   authors so well that we will be citing a lengthy passage.

   From [FBNorms]:

   "Facebook allows people to live-stream their suicide attempts "as
   long as they are engaging with viewers" but will remove footage "once
   there's no longer an opportunity to help the person". Pledges to kill
   oneself through hashtags or emoticons or those that specify a fixed

Elkins                   Expires April 20, 2019                [Page 11]

INTERNET DRAFT                  ifilter                 October 17, 2018

   date "more than five days" in the future shouldn't be treated as a
   high priority.

   These are tiny snippets from a cache of training materials that
   Facebook content moderators need to absorb, in just two weeks, before
   policing the world's largest social network.

   The guidelines also require moderators to learn the names and faces
   of more than 600 terrorist leaders, decide when a beheading video is
   newsworthy or celebratory, and allow Holocaust denial in all but four
   of the 16 countries where it's illegal - those where Facebook risks
   being sued or blocked for flouting local law.

   The documents detail what is and is not permitted on the platform,
   covering graphic violence, bullying, hate speech, sexual content,
   terrorism and self-harm. For the first time the public has a glimpse
   of the thought process behind some of the company's editorial
   judgements that go beyond the vague wording of its community
   standards or statements made in the wake of a live-streamed murder."

   The article goes on to posit that this may be the "most important
   editorial guide sheet the world has ever created".

   This use case brings up an issue which we may wish to consider. That
   is, there is no reason that Facebook, as a private company, needs to
   share with anyone what its methodology is for filtering. However,
   considering the enormous impact of Facebook, it is in the public
   interest to know the methodology.  In short, Facebook may be
   considered a public utility.

4.2 Blocking Extremist Activity

   From [BBCTECH], some of the content providers on the Internet are
   acting to censor content pertaining to potential extremist activity

   "In 2017, Google said it would dedicate more than 10,000 staff to
   rooting out violent extremist content on YouTube

   YouTube said staff had viewed nearly two million videos for violent
   extremism from June to December 2017

   YouTube said more than 98% of such material was flagged
   automatically, with more than 50% of the videos removed having fewer
   than 10 views

   Industry members have worked together since 2015 to create a database
   of "digital fingerprints" of previously identified content to better

Elkins                   Expires April 20, 2019                [Page 12]

INTERNET DRAFT                  ifilter                 October 17, 2018

   detect extremist material. As of December 2017, it contained more
   than 40,000 such "hashes"

   In 2017, Facebook claimed that 99% of all Islamic State and al Qaeda-
   related content was removed before users had flagged it. The social
   network said that 83% of the remaining content was identified and
   removed within an hour

   Between August 2015 and December 2017, Twitter said that it had
   suspended more than 1.2 million accounts in its fight to stop the
   spread of extremist propaganda. It said that 93% were flagged by
   internal tools, with 74% suspended before their first tweet."

4.3 Blocking Activity Inciting Violence

   [Myanmar] The United Nations report on the genocide of Rohingya
   muslims ties it to posts on Facebook.  Apparently, the Facebook
   content provider had very few people who could read Burmese.  So,
   posts were not reviewed.  The posts by the Myanmar military, intended
   to incite violence, indeed did so. There was wholesale killing of
   Rohingya muslims.  Facebook is now censoring such posts and has hired
   many Burmese speakers.

   [DailyStormer]   In August 2017 Cloudflare, one of the leading global
   CDNs, terminated the account of the Daily Stormer, a website
   advocating white supremacy and antisemitism, thus removing the
   website from the Internet. At the same time, several domain name
   registrars (GoDaddy, Tucows, Namecheap) discontinued the domain names
   used by the website. In the end, the website became accessible again
   by finding registries, registrars and hosters that would accept it,
   but in practice it was made almost unavailable for several weeks.

4.4 Copyright Protection

   Another reason for content filtering by OTTs, CDNs and hosting
   services is copyright protection.

   This has become a particularly active area since the EU adopted its
   digital copyright rules negotiating position (i.e., still in early
   stages) on 2018-09-12. Such rules will require all online platforms
   to implement automated content control at upload and screen the
   content for copyrighted material. [EU-DIGCOPY]

   We may wish to study how the music industry has evolved copyright
   protection over the past 100+ years in the US and elsewhere.

Elkins                   Expires April 20, 2019                [Page 13]

INTERNET DRAFT                  ifilter                 October 17, 2018

   In brief (US) they rely on designated third-party agencies (such as
   BMI, ASCAP, Harry J Fox) to provide licensing and collect royalties
   and distribute those back to copyright owners.  Statutory fees were
   set by the US congress. Private agreements are also possible, and
   common, of course.

   The music industry has developed a sophisticated ecosystem and rather
   than rely first on threats of criminal prosecution (which is possible
   in extreme cases) instead tries to convert as much of the problem as
   possible into civil claims (you used my work, you owe me money!).

   This is in stark contrast to the EU directive which approaches the
   problem via fines etc. and seems to create none of that

4.5 Filtering for Network and Computer Security

   Like ISPs, OTTs and CDNs also try to keep the network secure by
   making malicious or infected websites inaccessible. Search engines
   will mark results as unsafe; online platforms will disable links;
   hosting services and CDNs will terminate the web service.

   Some of the considerations in section 3.1 also apply here.  However,
   effective filtering measures at the Internet access point fully
   protect the end user.  To obtain the same effectiveness by acting at
   the core of the network, all the OTTs, hosting services and CDNs of
   the planet should be effective at taking down malicious content in a
   timely manner. Currently, this effectiveness varies; even a few
   "rogue" players being uncooperative to abuse and security takedown
   requests are enough to provide safe havens for attackers.

4.6 Content Filtering by End-Users

   Finally, the users themselves may want to block or mark content for
   several reasons. The content filtering types and purposes are the
   same described in section 3.2, but rather than relying on the ISP's
   infrastructure, they deploy appropriate software on their devices.
   This also includes user-controlled content classification mechanisms
   that avoid blocking content entirely, but still allow end users to
   preselect what they want to see or to miss on the Internet.

5 Security Considerations

   No new security vulnerabilities are introduced as a result of this

6  IANA Considerations

Elkins                   Expires April 20, 2019                [Page 14]

INTERNET DRAFT                  ifilter                 October 17, 2018

   No IANA actions are requested by this document.

6 References

6.1 Normative References

6.2 Informative References

   [Censorship] Hall, J., Aaron, M., Jones, B., Feamster, N., "A Survey
   of Worldwide Censorship Techniques",
   https://tools.ietf.org/html/draft-hall-censorship-tech-05, May 2018,

   [BBCTECH] https://www.bbc.com/news/technology-45495544, Sept. 2018

   [EU-DIGCOPY] http://www.europarl.europa.eu/news/en/press-
   copyright-rules, Sept. 2018

   [FBNorms] https://www.theguardian.com/news/2017/may/22/facebook-
   moderator-guidelines-extreme-content-analysis, May 2017

   accounts-myanmar-military-un-report-genocide-rohingya, August 2018

   [Child-Porn] https://gizmodo.com/fbis-disturbing-hacking-powers-
   challenged-in-court-over-1794885187, May 2017

   [UDHR] United Nations, "Universal Declaration of Human Rights", 1948,

   [EUCOPY] http://www.europarl.europa.eu/news/en/press-
   copyright-rules, September 2018

   [ITALY-REG] https://www.adm.gov.it/portale/lagenzia/monopoli-
   comunica/contrasto-illegalita, January 2007

   [UK-Controls] https://www.ispreview.co.uk/index.php/2017/10/uk-gov-
   October 2017

   [DailyStormer] Prince, M., "Why We Terminated Daily Stormer",
   https://blog.cloudflare.com/why-we-terminated-daily-stormer/, August

Elkins                   Expires April 20, 2019                [Page 15]

INTERNET DRAFT                  ifilter                 October 17, 2018

   [WikiOTT] WikiPedia, "Over-the-top media services",
   https://en.wikipedia.org/wiki/Over-the-top_media_services, October

Authors' Addresses

   Nalini Elkins
   Enterprise Data Center Operators (EDCO)
   EMail: nalini.elkins@e-dco.com

   Barry Shein
   Software Tool and Die
   EMail: bzs@theworld.com

   Vittorio Bertola
   Open Exchange
   EMail: vittorio.bertola@open-xchange.com

Elkins                   Expires April 20, 2019                [Page 16]

Html markup produced by rfcmarkup 1.129b, available from https://tools.ietf.org/tools/rfcmarkup/