ONI Releases 2009 Middle East & North Africa Research

By: Jillian C. York on 12 August 2009

The OpenNet Initiative is proud to announce the release of our Middle East and North Africa study. The 2008-2009 research can be accessed at http://opennet.net/research/regions/mena

Q: What are the filtering trends across the Middle East and North Africa?

While not all countries in the Middle East and North Africa filter the Internet, censorship across the region is on the rise, and the scope and depth of filtering are increasing. Testing has revealed political filtering to be the common denominator across the region; however, social filtering is on the rise. Many Arab countries have begun blocking explicit and morally objectionable content in the Arabic language that was previously accessible. While many regimes are transparent about social filtering, most continue to disguise political filtering practices by attempting to confuse users with different error messages.

Q: Are there any countries in the region that do not practice Internet filtering?

Based on ONI testing results, Algeria, Egypt, Iraq, Lebanon, and the West Bank do not currently filter any material; however, none of those are without regulations. Some Internet café operators in Lebanon, for example, have admitted to using surveillance software to monitor browsing habits of clients under the pretext of protecting security or preventing them from accessing pornography. Egypt has monitoring measures in place that require Internet café users to provide their names, email addresses, and phone numbers before using the Internet. Algeria holds ISPs legally responsible for sites they host.

Q: Which countries in the region filter most pervasively?

ONI evaluates each country based on the filtering of sites in four categories (political, social, conflict/security, and Internet tools), making it impossible to provide an overall rank. Rather, each country must be viewed according to the scope and types of content filtered.

Bahrain, Iran, Syria and Tunisia have the strictest political filtering practices in the region. The majority of ONI-tested countries heavily filtering social content are in the Middle East and North Africa and consist of Bahrain, Iran, Kuwait, Oman, Qatar, Saudi Arabia, Tunisia, and Yemen. A number of countries in the region also filter proxies and Internet tools.

Q: Have there been any changes in filtering since the 2007 study?

Overall, there has been an increase in filtering practices since 2007, and further measures to monitor Internet activities, particularly in Internet cafés, have been introduced. Additionally, countries that have been filtering political content continue to add more Web sites to their political blacklists.

Although increased filtering is the rule and unblocking the exception, there are a few instances of the latter since our last report. Syria has restored access to Arabic-language Wikipedia, Morocco has lifted a ban on a number of pro-Western Sahara independence Web sites, and Libya has begun to allow access to previously banned political sites. Additionally, Sudanese filtering of sites containing LGBT, dating, and health-related content has lessened since the last round of ONI testing.

Q: What content is typically filtered the most in the region? Are foreign media outlets accessible?

Political content is the type of content most frequently filtered in the Middle East and North Africa and includes (but is by no means limited to) sites related to political opposition parties or news sites and forums which espouse oppositional political views; occasionally such sites are unblocked following election periods. In those countries that filter social content, pornography, LGBT sites, and sites containing information on sexual health are often filtered.

Q: How significant is the filtering in Iran? Did censorship during the latest election increase?

Iran is among the strictest filtering regimes in the world, pervasively filtering political and social content, as well as Internet tools and proxies, and substantially filtering content related to conflict and security.

The presidential elections in 2009 led to an increase in online political organizing, which provided a further impetus for increasingly contentious controls on the Web sites used by legitimate opposition contenders. Facebook and YouTube were revealed by ONI testing to be blocked during the 2009 elections, despite being accessible shortly prior, and blogfa.com (a blogging platform) was found to be down during the election period.

The role of speech restrictions in Iran’s political realm are also evident in the guidelines passed down from the Supreme Council of the Cultural Revolution (SCRC) to the Committee in Charge of Determining Unauthorized Sites (CCDUS) in April 2009 that define allowable speech during the 2009 presidential elections for Web sites and ISPs. These guidelines outlined twenty categories of prohibited speech, including “disrupting national unity” and “creating negative feelings forwards the Islamic government.”

Q: The advent of social media has attracted the attention of filtering regimes. Which social media sites are filtered and where?

In the Middle East and North Africa, the filtering of social media and social networking sites has become relatively commonplace. For example, YouTube and Facebook are currently filtered Syria and Tunisia, and Orkut and Flickr are blocked in Iran and the UAE. Iran also filters a local social networking site, Balatarin.com, and the UAE and Saudi Arabia filter certain YouTube videos, though not the entire site.

Q: Is blocking used more frequently for Web sites in a local language compared to international Web sites?

Blocking Web sites in a local language is approximately twice as likely as blocking sites only available in English or other international languages. Many countries focus their filtering efforts on locally relevant Web content. Filtering in the Middle East tends to focus on local sites, but is often augmented with more international content.


Q: How long did you gather information?

Testing was conducted between 2008 and 2009, although the survey also draws upon work that ONI has conducted since 2003.

Q: How did you gather data?

ONI began by compiling large lists of Web sites – both globally and locally oriented – that cover a wide range of topics that have been the target of Internet filtering. The topics are organized into taxonomy of categories that have been subject to blocking, ranging from gambling, pornography, and crude humor to political satire and Web sites that document human rights abuses and corruption.

The tests were run within each country using software specifically designed for ONI. Where appropriate, the tests were conducted in multiple locations to capture differences in blocking behavior across Internet Service Providers (ISPs), and occasionally, across cities or regions. The tests were run over extended periods — multiple days or weeks — to distinguish active filtering from connectivity problems.

Q: Many of the countries surveyed are volatile locations with heavy government oversight. Understanding you had researchers on the ground in each country, how did you gather local data from within countries such as Iran, Syria, and Saudi Arabia?

ONI employs a unique multi-disciplinary approach that includes advanced technical means, using a suite of sophisticated network interrogation tools and metrics, and local knowledge expertise through a global network of regionally based researchers and experts. Many of our researchers are on the front lines, volunteering to conduct research worldwide.

Q: How many sites did you test?

We tested approximately 2,000 Web sites in each country, including a number of country-specific and locally-based sites for each.

Q: Why do countries censor Internet content?

The exact reasons a country chooses to filter content vary as much as the countries themselves, but overall, governments are increasingly seeking to regulate the Internet as they would any other activity within their states. In many respects, content controls are part of the maturing process of the Internet: Internet use is still largely unregulated, and as it grows, governments will likely formulate more laws and restrictions pertaining to online content and activity.

Filtering Techniques

Q: How do countries regulate content?

First and foremost, countries regulate content by legal and normative means, extending existing laws concerning defamation and licensing of traditional mass media to incorporate online content. These measures are designed to encourage self-regulation on the part of users and content producers. Countries also use technical means to limit access to content. These include IP blocking, DNS tampering, and proxy-based blocking methods. Each technique can be used at different levels of Internet access within a country. Internet filtering is most commonly implemented at two levels, at the Internet Service Provider (ISP) within the country and on the Internet backbone at the international gateway. These methods may overlap; an ISP may filter content using one particular technique while another technique is used at the international gateway. Pakistan is an example of a country that blocks at both the international gateway and at the ISP level.

Q: What equipment is needed to filter?

IP blocking is effective in blocking the intended Web site, and in most cases no new equipment needs to be purchased. It can be implemented in an instant; all the required technology and expertise is readily available. Depending on the network infrastructure within the country it may also be possible to block at or near the international gateways so that the blocking is uniform across ISPs. Countries new to filtering will generally implement IP blocking before moving on to more expensive filtering solutions. However, recent years have seen vendors move to fill this market niche, and filtering products are becoming increasingly commonplace. These products are designed for networks of all sizes, from personal use and Internet cafés to campus or corporate networks and ISP-level services.

ONI has identified several types of commercial products used by countries practicing pervasive backbone level filtering. The most common commercial software packages include SmartFilter (by Secure Computing and now owned by McAfee), WebSense, and Fortinet. Open source packages such as SquidGuard are also popular.

Q: What countries are examples of an “open Internet”?

The answer to this question is dependent on the definition of an “open Internet.” ONI tested for state sponsored filtering at the ISP level. In many countries, the majority of the population gets its access via Internet cafés or their place of work and study. These “edge locations” may in fact be filtered quite heavily – for example, an employer may block access to certain sites – so the absence of filtering at the ISP or backbone level is a relative measure of an “open Internet.” That said, our findings showed a significant number of countries with no evidence of filtering at the backbone level. In the Middle East and North Africa, Algeria, Egypt, Iraq, Israel, Lebanon and the West Bank are examples of such countries.

Q: Is a slippery-slope effect caused when a country begins to filter?

Our research suggests that, once the technical and administrative mechanisms are put into place, filtering tends to cover a relatively wide range of topics. The countries that target filtering on a small number of topics are in the minority. In addition, ONI research shows that several countries block content areas beyond those originally cited as the reason to begin Internet filtering.

Q: Are citizens aware of censorship in their countries?

In many countries, there is a widespread understanding that online censorship exists. In most places, the censor uses a “block page” that tells the user that the site they are attempting to access is banned. News about blocked services may travel quickly via word-of-mouth or traditional media, particularly in the case of a popular service such as YouTube. In other cases, where filtering is pervasive and may occur at the search engine level as well, citizens are faced with the “1984 problem,” where they may not realize the extent of the information the government is preventing them from accessing, as they have no way of knowing it exists in the first place.

Q: Can individuals take action to have content blocks reversed?

A small group of countries, including Saudi Arabia, Oman and the United Arab Emirates, and some ISPs in Iran, allow Internet users to write to authorities to register a complaint that a given Web site has been blocked erroneously. In other cases, groups of concerned citizens or other informal advisory bodies have quietly reasoned with authorities to reverse decisions to block specific sites. Most countries, however, provide no mechanism for appealing the decision to block a Web site.

Q: How often are sites filtered?

Modifications can be made to the blocking efforts of a country by the authorities at any time. Sites can be added or removed at their discretion. For example, during tests in Iran the Web site of The New York Times was blocked for a single day. China is known for frequently blocking and unblocking certain social media sites such as YouTube and Twitter. Some countries, such as Moldova, have also been suspected of introducing temporary filtering around key events such as elections.

Q: Are there technologies to get around censorship?

There is a growing set of tools that enable people to get around online censorship. One of the most common types of these tools is the anonymizer, which accesses the Internet while hiding information that could be used to identify the user. One common anonymizer is Tor.

Another tool used to circumvent filtering is Psiphon, which allows users to access blocked sites in countries where the Internet is censored by turning a regular home computer into a personal, encrypted server capable of retrieving and displaying Web pages anywhere. Psiphon relies on a trusted network to circumvent Internet filtering.

These tools are increasingly used globally.


Q: What is the OpenNet Initiative?

The OpenNet Initiative is a partnership between the Citizen Lab at the Munk Center for International Studies at the University of Toronto, the Berkman Center for Internet & Society at Harvard Law School, the Oxford Internet Institute at the University of Oxford, and the Advanced Network Research Group at the Cambridge Security Programme at Cambridge University. The partnership seeks to investigate and document state filtration and surveillance to promote and inform wider public dialogs about such practices.

Q: Where can I find additional information about the survey?

The ONI Web site has interactive maps, country profiles and regional overviews, as well as individual reports covering specific instances of blocking. Our Web site is located at www.opennet.net.

Q: What can Western companies do about the increasing amount of censorship?

Western companies are faced with the challenge of deciding how to do business in regimes that censor and carry out filtering on the Internet. As the number of such states is growing, this problem is besetting virtually all global companies in the Internet and telecommunications space. The Global Network Initiative (http://www.globalnetworkinitiative.org/), a multi-stakeholder group of companies, civil society organizations (including human rights and press freedom groups), investors and academics, has created a collaborative approach to protect and advance freedom of expression and privacy in the ICT sector.

Q: What is next for ONI?

An in-depth analysis of our 2008-2009 results will be published in a volume entitled Access Controlled due for release in early 2010 as a follow-up to our 2008 volume, Access Denied: The Practice and Policy of Global Internet Filtering. We will follow up the release of our Middle East and North Africa survey with timed releases of results from Asia, Latin America, Europe, the Commonwealth of Independent States, Sub-Saharan Africa, Australia and the US/Canada.