Global Internet filtering in 2012 at a glance

By: ONI Team on 3 April 2012
Posted in Political filtering, Social filtering, Conflict and security filtering, Internet tools filtering

The OpenNet Initiative (ONI) has been documenting Internet filtering globally since 2003. Since that time, the number of countries found to be engaging in the censorship of online content has increased dramatically. While early reports of Internet filtering only identified a small handful of authoritarian countries engaged in the practice, the number has since grown to the extent that the practice can now be considered a pervasive global norm.

When we last reported on the prevalence of Internet filtering globally in early 2010, we estimated there were over 500 million Internet users residing in countries that engage in the systematic filtering of online content. In 2012, after close to a decade of documenting Internet censorship worldwide, ONI estimates that number has increased to over 620 million.1

Since 2003, ONI has conducted testing of Internet filtering in 74 countries. (See here for ONI’s testing methodology and here for raw summarized results of ONI testing results.) Of these 74 countries, 42 have been found to engage in some form of filtering of content that ONI tests for, while 21 have been found to be engaging in “substantial”2 or “pervasive” filtering based on the breadth and/or depth of content filtered.3 (Note: These results are based on testing conducted over different time periods and thus represent a snapshot of accessibility in time for each country. See ONI’s summarized results for further information about testing periods.) Within those countries found to be engaging in “substantial” or “pervasive” blocking, ONI has estimated that the number of Internet users is approximately 620 million, which makes up 31% of all Internet users worldwide.4 When this is widened to include countries that engage in “selective” blocking, the number of affected users grows to over 960 million, or 47% of all Internet users.

While this number is only a rough estimate based on our testing sample and available data, it does reflect the steady increase of Internet blocking seen worldwide, a trend that shows no signs of abating. While there have been several notable examples of countries reducing the depth and/or scope of filtered content (see Burma and Tunisia post-Arab Spring), there have also been others that have begun or given signs they will begin filtering (see Egypt and Indonesia). It is also important to note that the growing number of users in countries that practice filtering also reflects the rapid increase in the number of Internet users in the global South and East, regions in which stricter information controls are often the norm.

The ONI has also observed information controls beyond Internet filtering. For example, in the past year we have observed a growing number of incidents where states have disrupted or tampered with communication networks for political purposes, including around elections and public demonstrations. The ONI calls this practice just-in-time-blocking—a phenomenon in which access to information is denied during important political moments when the content may have the greatest potential impact, such as elections, protests, or anniversaries of social unrest. In 2011 both Egypt and Libya severed all Internet access for brief periods of time during the “Arab Spring.” Similar tactics have been employed in Nepal (2005), Burma (2007), and China (2009). Just-in-time blocking in response to protest and social unrest has also been called for, and in some cases implemented, in democratic states. In response to the 2011 UK riots the government has put forth proposals to restrict citizens from using social media “when we know they are plotting violence, disorder and criminality.” In the summer of 2011 the Bay Area Rapid Transit System (BART) shut down cell phone service to four stations in San Francisco in an effort to disrupt the organization of planned protests.

Continuing increases in Internet censorship and the widening adoption of other information control have troubling implications for the future of free expression and information sharing online.

  • 1. Several important caveats about these figures:
    • Due to legal restrictions, ONI does not test for the filtering of material related to child exploitation. Countries which are believed to exclusively filter content relating to child exploitation are excluded from these estimates.
    • These results are based on ONI testing in 74 countries and thus likely underestimate the total number of countries engaged in Internet filtering worldwide.
    • ONI does not conduct testing on institutional or private networks (for example, in schools or on internal corporate networks) and thus these figures underestimate the total breadth and impact of filtering globally.
  • 2. See here [pdf] for a breakdown of how ONI defines “selective”, “substantial” and “pervasive” filtering, as well as a list of the types of content that ONI tests for.
  • 3. Countries classified by ONI as engaging in “substantial” or “pervasive” filtering of one or more content category are Armenia, Bahrain, China, Ethiopia, Gaza and the West Bank, Indonesia, Iran, Kuwait, Burma/Myanmar, Oman, Pakistan, Qatar, Saudi Arabia, South Korea, Sudan, Syria, Turkmenistan, United Arab Emirates, Uzbekistan, Vietnam and Yemen.
  • 4. Estimates of Internet users are based on data from the World Bank World Development Indicators, which are themselves drawn from data from the International Telecommunications Union.