Posted 31 March, 2011 by Christopher Wilson

Exporting Censorship Technology

The Open Net Initiative released a report this week, documenting how middle eastern countries are censoring the web for national users.  That this is happening is old news. A number of press freedom indexes have been charting this for some time, including the annual list of “Internet Enemies” that Reporters Without Borders has been putting out since at least 2007. An abundance workarounds have also been developed, both in terms of national information aggregators such as, and online DIY libraries (Sesawe has a broad collection for both circumvention and safety) .

This report makes an important contribution however, by identifying the “who” in detail (Netsweeper & Websense) and “where”, which is important for national activists. What is even more compelling is the “how”, and the ONI title makes it pretty clear: West Censoring East: The Use of Western Technologies by Middle East Censors, 2010-2011. Of course, we have also known about this dynamic in general for some time, and there is the occasional revelation regarding specific relationships, such as the post-revolution press on how European and American Companies supported digital censorship by the Mubarak regime. But it is the details in this report that are particularly useful and devlish.

Most striking is the “out-of-box” nature of solutions that seem to be favored by the repressively-inclined.  The report documents ready for use software solutions that allow censors to choose what kind of content they wish to block from a checklist (“skeptical views of Islam, secular and atheist discourse, sex, GLBT, dating services, and proxy and anonymity tools,”, including web content offensive to “social, religious or political ideals,” as one company’s sales pitch has it. This is problematic for a number of reasons.  There is the obvious and awkward turpitude of western corporate support to suppression in the developing world, and there is more to be said about that below.

Another problem is the indiscriminate grouping of presumably legitimate censorship (read child porn), and political dissidence.  While this makes sense from a technical perspective that operates above actual content, it seems as though the obvious potential for cretinous application should mandate a more responsible approach.  Despite the dubitable impact of voluntary compacts have had in promoting free expression to date (see Google’s latest commitment here), self regulation strikes me as the only logical intervention here, both for technical reasons and market reasons. This type of work strikes me as a clear job for the Business and Human Rights Community, who should probably get engaged sooner than later, to shape the discourse and take advantage of popular sentiment on the issue.

While the issue’s importance will likely only grow in parallel with global internet penetration rates and widespread appreciation for the role that digital tools can play in processes of democraticzation, it seems likely that that importance will become less obvious as censorship tools and techniques become more opaque–another reason to engage now, while the engaging is good.

A third problem is the tools’ failure to accurately identify content that should be blocked. The ONI report identifies a number of reasons why web filter technology blocks content not intended for censure, including reliance on large databases maintained by the service provider, human error and over reliance on keywords.  This has produced some amusing results, such as the temporary blockage of George Bush’s official page for some American markets in 2007, presumably due to his last name. It also leads directly to a fourth problem–that these technologies are easily manipulated by third parties. Jillian York (one of the report’s authors) has blogged about how easy it is for anyone at all to manipulate this software through comment spamming, in order to get a web site blocked. Given the lack of oversight described in the ONI report, and recent descriptions of how individuals have worked strategically to take down activists’ facebook pages, this is especially troubling in the Middle Eastern context, where revolution can move quickly to dramatic sectarianism, as we have seen in Egypt.

Perhaps most troubling, is the sinking feeling upon finishing the report, that companies are generally, actively and knowingly selling tools of repression to autocrats.  This is not clearly true of both companies covered in the ONI report, but as further revelations emerge from a fragile and post-revolutionary Egypt (and it remains to be seen what gems the treasure trove of Amn al Dawla will yield), there is reason to be pessimistic.  The work being done by ONI, Sesawe and others is critical in this regard, in as far as it makes the issue an issue for people in the countries where these companies are based, and this might be strengthened by strategic engagement in the business and human rights community . But as autocrats get better at sharing digital skills, and in the face of a widespread belief in the Arab world that the internet should be censored, it is clear that more is needed. In the end, this will of course be work done on the ground by the people to whom it matters most, and who can do it best.  But as we have seen in the satellite television sector, regional actors can have a dramatic impact on local understandings of the role of media, and as donors contemplate the future of their MENA portfolios, it is hard not to think that Egypt has an important role to play here.

1 thought on “Exporting Censorship Technology”

[…] a national download ban on “copyrighted content” on the same day it denounced the export of censorship technology to […]

Leave a Reply

Your email address will not be published. Required fields are marked *

Related articles