tial Security Council. As a result, pressure was exerted on UzSCINET to cooperate with
authorities, and over time the network became a ‘‘testing ground’’ that security forces
used to develop a system for selecting and blocking unwanted Web sites. As late as
2005, the system was far from comprehensive, with previous ONI research showing a
great deal of divergence among the various ISPs—some comprehensively blocked content, while others allowed unfettered access. The suspicion is that some commercial
ISPs had close connections with President Karimov’s inner circle and hence were able
to withstand pressure to implement filtering, which gave them a commercial advantage (as users who wished to access such content would pay to access the Internet
through these ISPs).
Uzbekistan’s principal intelligence agency, the National Security Service (SNB), monitors the Uzbek sector of the Internet and ‘‘stimulates’’ ISPs and Internet cafe ´s to practice self-censorship. Soviet-style censorship structures were replaced by ‘‘monitoring
sections’’ that work under SNB’s guidance. There is no mandatory government prepublication review, but ISPs risk having their licenses revoked if they post ‘‘inappropriate’’
information. Occasionally, the SNB orders ISPs to block access to opposition or religious Web sites. 43 A survey of internet filtering practices among Uzbek ISPs was conducted by ONI in January 2007. Respondents confirmed that they use filtering
applications including SquidGuard and FortiGuard. The SNB’s censorship is selective
and often targets articles on government corruption, violations of human rights, and
organized crime. Usually, it affects URL-specific pages instead of top-level domain
names. Uzbek ISPs block entire Web sites or individual pages upon SNB’s unofficial
requests. Accessing a blocked page redirects the user to a search engine or to an error
message such as ‘‘You are not authorized to view this page.’’ The retransmission of
blocked channels is also prohibited.
The SNB regularly exchanges data with Russian intelligence sources and allegedly
collaborates with the Russian Foreign Intelligence Academy. The SNB also utilizes a
blacklist and keyword approach. The SNB practice of active surveillance contributes to
self-censorship among Internet operators and the Internet community as a whole.
Most users will not engage in topics that touch on unpopular government policies
relating to human rights in the country out of fear of arbitrary prosecution by the
authorities.
ONI Testing Results
In 2007 and 2008, the OpenNet Initiative conducted testing on five main ISPs in the
country: ArsInform, Buzton, Sharq Telecom, Sarkor Telecom, and Uzbektelecom. The
test results show pervasive blocking of different categories of Internet content, including local and international human rights sites (inter alia, content promoting the rights
of journalists working in repressive regimes), local and regional media sites, opposition
Uzbekistan
271
authorities, and over time the network became a ‘‘testing ground’’ that security forces
used to develop a system for selecting and blocking unwanted Web sites. As late as
2005, the system was far from comprehensive, with previous ONI research showing a
great deal of divergence among the various ISPs—some comprehensively blocked content, while others allowed unfettered access. The suspicion is that some commercial
ISPs had close connections with President Karimov’s inner circle and hence were able
to withstand pressure to implement filtering, which gave them a commercial advantage (as users who wished to access such content would pay to access the Internet
through these ISPs).
Uzbekistan’s principal intelligence agency, the National Security Service (SNB), monitors the Uzbek sector of the Internet and ‘‘stimulates’’ ISPs and Internet cafe ´s to practice self-censorship. Soviet-style censorship structures were replaced by ‘‘monitoring
sections’’ that work under SNB’s guidance. There is no mandatory government prepublication review, but ISPs risk having their licenses revoked if they post ‘‘inappropriate’’
information. Occasionally, the SNB orders ISPs to block access to opposition or religious Web sites. 43 A survey of internet filtering practices among Uzbek ISPs was conducted by ONI in January 2007. Respondents confirmed that they use filtering
applications including SquidGuard and FortiGuard. The SNB’s censorship is selective
and often targets articles on government corruption, violations of human rights, and
organized crime. Usually, it affects URL-specific pages instead of top-level domain
names. Uzbek ISPs block entire Web sites or individual pages upon SNB’s unofficial
requests. Accessing a blocked page redirects the user to a search engine or to an error
message such as ‘‘You are not authorized to view this page.’’ The retransmission of
blocked channels is also prohibited.
The SNB regularly exchanges data with Russian intelligence sources and allegedly
collaborates with the Russian Foreign Intelligence Academy. The SNB also utilizes a
blacklist and keyword approach. The SNB practice of active surveillance contributes to
self-censorship among Internet operators and the Internet community as a whole.
Most users will not engage in topics that touch on unpopular government policies
relating to human rights in the country out of fear of arbitrary prosecution by the
authorities.
ONI Testing Results
In 2007 and 2008, the OpenNet Initiative conducted testing on five main ISPs in the
country: ArsInform, Buzton, Sharq Telecom, Sarkor Telecom, and Uzbektelecom. The
test results show pervasive blocking of different categories of Internet content, including local and international human rights sites (inter alia, content promoting the rights
of journalists working in repressive regimes), local and regional media sites, opposition
Uzbekistan
271