Blogs

The rise of the European upload filter

By EDRi · June 20, 2012

This article is also available in:
Deutsch: [Vom Aufstieg der Upload-Filter in Europa | https://www.unwatched.org/EDRigram_10.12_Vom_Aufstieg_der_Upload-Filter_in_Europa?pk_campaign=edri&pk_kwd=20120620]

In 2011, the European Union decided against the introduction of
mandatory filtering in Europe, because a democratic analysis of the
evidence showed that this was not necessary. In 2012, the European Union
is working on a variety of projects to introduce “voluntary” upload
filters and, because they would be introduced on a so-called “voluntary”
basis, there will be no democratic analysis.

There are at least three different initiatives currently underway with
this target. Firstly, Commissioner Kroes’ “CEO Coalition”. As previously
mentioned in the EDRi-gram, this initiative involves the Commission
inviting big business to propose measures that will make the Internet a
“safer place for children”. The fact that Facebook is in charge of
privacy settings and Microsoft is in charge of “notice and takedown”
appears to raise no concerns about the process being instrumentalised
for business purposes.

Microsoft’s history in “notice and takedown” is hardly exemplary. Recent
cases showing that Microsoft, using Google’s global application of US
copyright law, has repeatedly demanded that Google removes links from
its search results – links which remain available on Microsoft’s own
search engine. Microsoft, on the other hand, has also developed a
product called “photoDNA,” used by Facebook UK as an upload filter to
prevent known child abuse material being added to their site. What
happens when somebody tries to do this? Nobody knows. What happens if a
criminal tries to upload innocent parts of images as a way of filtering
his/her own collection of illegal images to identify images unknown to
the police? We don’t know. How big is the risk that this could lead to
incentives to creating new illegal images and new abuse? We don’t know.

In the same vein, the “Safer Internet Programme 2012” announced that one
of its strategies was to “develop a pilot test for trusted hash code /
fingerprint series for preventing re-uploading of identified child
sexual abuse material”. Remarkably, the Commission’s proposal includes
an explicit reference to PhotoDNA, Microsoft’s product. The plan is also
to extend beyond photographic material and cover videos.

Finally, the European Parliament is currently working on a
non-legislative resolution on “the protection of minors in the digital
world,” with Silvia Costa (S&D, Italy) as MEP in charge. Neither her
draft report nor any of the amendments make a reference to filtering of
any kind. However, the text, largely (as it seems reasonable to assume)
donated by the child protection industry, is replete with references to
“preventing” various online activities, without any discussions in the
Parliament about filtering uploads to the Internet. The draft report
also masterfully confuses illegal content with unspecified “unsuitable”
content arguing that “measures to prevent illegal online content lead to
differing approaches to the prevention of unsuitable conduct”.

But one can’t really oppose the use of such a far-reaching strategy for
child protection. If it protects children, why ask questions about
whether there will be unforeseen or even counter-productive
consequences? Why ask questions about whether this is the most effective
use of resources? Why ask for democratic oversight or evidence of
usefulness? And, of course, nobody would be so cynical as to re-use the
technology for any other purpose, would they?

Well, apart from terrorism, of course. The EU-funded “Clean IT” (not to
be confused with the Iranian “clean Internet”) project is proposing the
creation of a database of content “considered” illegal. “Why not try and
create a database where internet companies can check it to see if it’s
known illegal material” asks the project manager? And if it protects
society, why ask questions about whether there will be unforeseen or
even counter-productive consequences? And all of the questions that are
not asked about upload filters for child abuse material are not asked
again in relation to terrorism.

Well, apart from copyright, of course. The European Commission’s report
from 2010 on the application of the IPR Enforcement Directive argues
that, “given intermediaries’ favourable position to contribute to the
prevention and termination of online infringements, the Commission could
explore how to involve them more closely.”

Safer Internet Programme
http://ec.europa.eu/information_society/activities/sip/docs/call_2012/work_programme_final.pdf

IPR Enforcement Directive
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2010:0779:FIN:EN:PDF

Draft European Parliament report on protection of minors (2.04.2012)
http://www.europarl.europa.eu/sides/getDoc.do?type=COMPARL&mode=XML&language=EN&reference=PE486.198

Clean IT project considers terrorist content database (6.06.2012)
http://www.itnews.com.au/News/303729,clean-it-project-considers-terrorist-content-database.aspx

Busted: Microsoft Harbors BitTorrent Pirates (27.05.2012)
http://torrentfreak.com/busted-microsoft-harbors-bittorrent-pirates-120527/

Facebook & PhotoDNA (19.05.2011)
http://blogs.technet.com/b/microsoft_on_the_issues/archive/2011/05/19/facebook-to-use-microsoft-s-photodna-technology-to-combat-child-exploitation.aspx

(Contribution by Joe McNamee – EDRi)