Internet Filtering in Vietnam in 2005-2006: A Country Study

1. Executive Summary
2. Political, Technical, and Legal Context in Vietnam

A. Sensitive / Controversial Topics for Media Coverage
B. Internet Infrastructure and Access
C. Legal Background

3. Testing Methodology

A. Methods
B. Results Analysis
C. Methods Specific to Vietnam
D. Topics Tested

4. Results and Analysis

A. Summary
B. Global List Results
C. Vietnam-Specific Resultsstudies/vietnam/
D. Filtering Methods

5. Conclusion

Appendix 1
Appendix 2
Appendix 3
Appendix 4
Appendix 5


1. EXECUTIVE SUMMARY (1)


The Socialist Republic of Vietnam regulates access to the Internet by its citizens extensively, through both technical and legal means. This study by the OpenNet Initiative (ONI) finds that the Vietnamese state attempts to block citizens from accessing political and religious material deemed to be subversive along various axes. The technical sophistication, breadth, and effectiveness of Vietnam's filtering are increasing with time, and are augmented by an ever-expanding set of legal regulations and prohibitions that govern on-line activity. Vietnam purports to prevent access to Internet sites primarily to safeguard against obscene or sexually explicit content. However, the state's actual motives are far more pragmatic: while it does not block any of the pornographic sites ONI tested, it filters a significant fraction - in some cases, the great majority - of sites with politically or religiously sensitive material that could undermine Vietnam's one-party system. Vietnam's Internet infrastructure and market are dynamic and fast-changing, but it seems inescapable that the state's on-line information control will deepen and grow.

Vietnam focuses particular effort on blocking access to sites related to topics that challenge the state's political orthodoxy, such as those treating political dissidents, political democracy, or the proposed Vietnam Human Rights Act in the United States Congress. Sites on topics related to domestic religious faiths, such as Buddhism and Caodai, are also subject to blocking, though less extensively. In nearly all cases, sites in the Vietnamese language are far more likely to be blocked than sites in the English or French languages.

Like many states in Southeast Asia and in other regions of the world, Vietnam tries to leverage the Internet to provide economic development and benefit, while simultaneously struggling to limit access to content that might destabilize the Communist state.2 This tenuous balancing act finds its clearest expression in Vietnam's 2001 decree on management, provision, and use of Internet resources, which requires that Internet "management capacity must be in line with development requirements, while at the same time consistent measure must be taken to prevent abuse of the Internet to affect the national security and break national virtues and traditional good customs" to achieve the "national cause of industrialization and modernization."3 Among other methods, the state monitors the use of cybercafés and employs filtering of Internet content to control the information its citizens can access online.

Vietnam's Internet filtering regime demonstrates an ongoing struggle between the desire to increase Internet usage for economic purposes and the desire to limit its effects politically, and between the capabilities of filtering technology and the limitations of bandwidth, ISP resources, and users' tolerance. This system, more than any other that ONI has studied, shows dramatic change over time and bears careful monitoring of its development.


2. POLITICAL, TECHNICAL, AND LEGAL CONTEXT IN VIETNAM

Internet filtering operates within a broader context of legal regulation, technical architecture, and political and social norms. This section explores these areas and their relevance to Vietnam's Internet filtering regime.


A. Sensitive / Controversial Topics for Media Coverage

Sensitive issues in Vietnam include corruption, ethnic unrest, and political opposition, including topics related to democracy or criticism of the state's single-party system.4 The Vietnamese government is concerned about allegations of corruption after several high-profile incidents, including the firing of the minister for agriculture and rural development in June 2004 for failing to prevent a multi-million dollar (USD) corruption scandal in his ministry,5 allegations of corruption in the Vietnam Post and Telecommunications Corporation (VNPT, the state telco provider),6) and embezzlement in the Transportation Ministry that led to the minister's resignation and to the arrest of a deputy minister.7 The Prime Minister announced plans for an anti-corruption agency in late 2004.8

The state is also concerned about ethnic unrest, particularly in rural areas. For example, Vietnamese authorities put down a disturbance of thousands of Montagnards (members of a minority ethnic group who are primarily Christians) in Gia Lai province in April 2004.9 The government then forced a Webmaster to close a site after he copied an article about the violent protests; the student who ran the site paid a fine and had to write a letter of self-criticism.10

Pornography, and other sexually explicit content, is at least nominally a concern for state regulators. The head of a provincial Post and Telematics Office cited statistics that 90% of young Vietnamese Internet users had accessed pornography, and 26 of 28 cybercafés checked by officials had been used to access sexually-oriented material.11 The Deputy Minister of the Ministry of Culture and Information noted his concern about the spread of this content to new forms of communication, including "sex chat, sex films on mobile phones, etc."12 He stated that the Internet "can also be a place where erroneous or harmful news content [and] pornographic materials [are available]? If online information is not controlled well, it can be a bad influence." 13

Vietnam worries particularly about the use of the Internet to promote political reform, opposition groups, or viewpoints critical of the state in areas such as international relations.14 Critiques of the state or its high-level leaders remain taboo.15 Amnesty International claims that at least 10 people have been arrested for political activity using the Internet, with seven sentenced to prison.16 For example, Nguyen Dan Que was sentenced to 30 months in jail for writing and posting on the Internet an article about the lack of freedom of information in Vietnam; he was released as part of an amnesty in February 2005, but remains under surveillance.17 Dissident Le Chi Quang was sentenced to four years in prison for posting an online article that criticized a border accord between China and Vietnam.18

Finally, Vietnam is sensitive to criticism regarding its record on human rights. For example, a blogger in Vietnam posted a photo of a copy of the Asia Wall Street Journal purchased in Ho Chi Minh City.19 The newspaper had been edited using old-fashioned means: the offending text was blacked out. The obscured text noted widespread criticism of Vietnam's human rights record and the state's designation as a "country of particular concern" in the U.S. State Department's annual report on religious freedoms worldwide.


B. Internet Infrastructure and Access

Vietnam's Internet system is growing rapidly. Change is constant, and it is difficult to describe the situation "on the ground" with complete accuracy; indeed, even knowledgeable officials who shared information with ONI often disagreed even about basic issues such as the number of ISPs actually operating. The growth in user base is clear. In 2004, Vietnam estimated it had 5.34 million Internet users, including 1.44 million Internet subscribers;20 by October 2005, Vietnam reportedly had 9.2 million users, including 2.5 million subscribers, creating an Internet penetration rate of 11.1%.21 Public, rather than home, Internet access is the norm in Vietnam. Most users access the Internet from cybercafés, making control over this venue important for the state.22 Cybercafé access is relatively inexpensive: in March 2006, for example, using a computer with broadband Internet access for one hour in cafés in Ho Chi Minh City cost roughly 6000 VND (approximately $0.40 USD). It is common for businesses such as restaurants, art galleries, and bars to offer Internet access for a fee. Many hotels advertise Internet access as well. 23

In addition, because the Vietnamese telecommunications system is patterned on the French system, postal offices are important providers of services such as telephone calls (PSTN) and Internet access. Voice over Internet Protocol (VoIP) is an increasingly popular means of communication; cybercafés in Ho Chi Minh City, for example, often explicitly advertise VoIP services, and some only provide VoIP (rather than a range of Internet services). Internet access via satellite is not permitted, in part to maintain the monopoly on satellite services enjoyed by VNPT. Pre-paid Internet access cards are readily available and provide a measure of anonymity to users who employ them.23 The state expressly works to reduce access costs "to quickly popularize the Internet in Vietnam and further enhance the competitive edge of Vietnamese enterprises in the process of international economic integration."24

Vietnamese online services, though, are still nascent: a research group from the Ho Chi Minh City University of Natural Sciences is working to create a Vietnamese search engine based on keywords, as there are few search engines currently available to Vietnamese users.25 Vietnam has adopted TCVN 6909:2001 as the official method of representing Vietnamese 16-bit characters online,26 though older character sets such as VIQR and Windows Vietnam cp1258 may also be used.27

Vietnam controls the allocation of domain names under the country code top-level domain (ccTLD) .VN through the Vietnam Internet Center (VNNIC).28 As of mid-July 2005, the .VN domain reportedly had 11,533 registered domains, with about 400-600 new .VN domain names being registered each day.29 Vietnam is also planning to implement a state-controlled Vietnamese-language second-level domain,30 presumably similar in concept to the new Chinese-language top-level domains set up by China.31 Blogging remains quite limited in Vietnam, and most bloggers who post from inside the state do so in English. Informal discussions with bloggers in Vietnam indicate that some users are concerned about repercussions if they post in Vietnamese to a blog, even about less sensitive topics.

State regulation determines how Internet connectivity in Vietnam is organized. At the edge of the network, Internet agents, such as cybercafés, connect to the Internet service provider (ISP) with which they have signed their agency contract.32 ISPs may connect with each other and with Internet exchange points (IXPs), except that private ISPs may not connect with each other in peering arrangements.33 Online service providers (OSPs) and Internet content providers (ICPs) may connect to ISPs and IXPs. 34 IXPs can connect with each other (peering) and with the international Internet.35 Thus, connectivity is managed so that only IXPs may maintain connections with the wider Internet.36 ISPs must obtain upstream access by leasing it from IXPs. This likely makes filtering Internet content easier for Vietnam by limiting external access points that must be controlled. In the past, all IXPs were required to purchase physical connectivity from Vietnam Telecom International (VTI). Currently, IXPs can theoretically maintain independent connections to the international Internet, but it is not clear how many do so in practice. Foreign-owned entities that can purchase VSAT connections can obtain unfiltered Internet access, but this is relatively rare in Vietnam, and largely irrelevant for Vietnamese Internet users. Vietnam's network topology is in constant flux; our analysis applies to the infrastructure as it existed at the time of our testing.

Vietnam's primary, state-owned gateway, Viet Nam Data Communications Co. (VDC),37 is owned by Vietnam Posts and Telecommunications Group.38 (VDC was previously owned by the Vietnam Posts and Telecommunications Corporation, or VNPT, which also became part of the Vietnam Post and Telecommunications Group after a reorganization approved in March 2005 and implemented in January 2006.39 The reorganization appears to contemplate splitting VNPT into three different Telecommunications Corporations divided by geographic region.40 The Group is now known as VNPT.41) A Ministry of Posts and Telecommunications official stated that there are three international gateways, with VNPT / VDC, Viettel, and FPT each controlling one.42Officials at those companies, however, indicate that in practice VNPT / VDC has three gateways, Viettel has three, and FPT has four.43

Formally, the Ministry of Posts and Telecommunications lists 13 Internet access services providers (IASPs), six IXPs, and 11 OSPs as licensed in Vietnam.44 However, ONI research indicates that entities with the legal right to offer services in Vietnam may not be yet doing so, and knowledgeable sources disagree about the precise number of ISPs actually operating.

Type of Entity Company
Internet Access Service Provider (IASP) VNPT
FPT
Viettel Corp.
Saigon Posts and Telecommunications Service Corp. (SPT) [Saigon Net]
Hanoi Telecom Co.
Internet Service Stock Co. (OIC)
TIE Trade Import Export Co. (TIENET)
ELINCO
TECHCOM
Quality Technology Network Corp. (QTNET)
Thanh Tam Stock Co.
Netnam Co.
Viet Khang Co. (XVNET)
Internet Exchange Providers (IXP) VNPT45
FPT
Viettel
SPT
Hanoi Telecom Co.
Electric Telecommunication Co.46
Online Service Provider (OSP) VNPT
FPT
Viettel
SPT
Hanoi Telecom
OIC
TIENET
ELINCO
TECHCOM
QTNET
Thanh Tam

Figure 1 ? Licensed Internet Entities

In reality, only six ISPs have sufficient market share to be considered viable. Indeed, VNNIC permits only eight entities to operate under the .VN domain, and only seven IASPs offer access to subscribers.47 The ISPs with significant market share are listed in the chart below.

ISP URL Market Share48
VNPT http://www.vnpt.com.vn/ 44.34%
Vietnamese Corporation for Financing and Promoting Technology (FPT) http://www.fpt.vn/
http://www.hcm.fpt.vn/
28.62%
Viettel http://www.vietel.com.vn/ 13.67%
Saigon Net http://www.saigonnet.vn/
http://www.saigonpostel.com.vn/
6.44%
Netnam http://www.netnam.vn/
http://www.netnam.vn/services.htm
5.16%

Figure 2 ? List of Major Vietnam ISPs

ONI was able to obtain data on how certain of Vietnam's ISPs obtain international Internet access:

ISP International Access (Upstream) Provider(s)
VDC / VNPT Singtel, Kornet, Chunghwa Telecom, Malaysia Telecom, NTT, Reach, KDD, ChinaNet, Fusion
Viettel Hutchinson, Dacom, SingTel, VSAT
FPT Hutchinson, Reach
ETC Kornet
SPT Reach

Figure 3 ? Upstream Access Providers to Vietnamese ISPs

The cost of upstream bandwidth is considerable for Vietnamese ISPs, accounting for as much as 80% of their input costs. Moreover, bandwidth may not be reliable; the deputy minister of the MCI admitted that some domestic Web content providers host their sites on servers abroad because they "have transmission lines of higher quality."49 In total, Vietnam leases roughly 2200 Mbps of international line capacity. This international leased line capacity is allocated as follows:

ISP Share of International Leased Line Bandwidth
VNPT 76.32%
FPT 13.96%
Viettel 9.46%
SPT .18%
ETC .09%

Figure 4 ? ISP Shares of International Bandwidth

All ISPs in Vietnam are either state-owned enterprises or entities that have been "equitized" (a process where a state-owned enterprise issues shares to the state, to management, and to workers, but not to the private market), except for OCI, which is privately owned.50 For example, FPT was formerly a state-owned enterprise, but was equitized. ONI research indicates that state-owned, equitized, and private ISPs face the same fees ? there is no internal discrimination among ISPs by the state. While any Vietnamese firm can operate as an ISP, only companies that are state-owned can be IASPs or IXPs.51 Majority ownership of telecommunications entities in Vietnam by foreign investors is not permitted. This limitation makes control over Internet content easier to achieve for Vietnam.


C. Legal Background

1. General Media Regulation

Like most Communist countries, Vietnam nominally guarantees freedom of speech, of the press, and of assembly through constitutional provisions, but state security laws and other regulations trump or eliminate these formal protections (commonly referred to as the "national security override"). Thus, Article 69 of the Constitution of the Socialist Republic of Vietnam states that a "citizen shall enjoy freedom of opinion and speech, freedom of the press, the right to be informed, and the right to assemble, form associations and hold demonstrations in accord with provisions of the law."52 However, media in Vietnam are tightly controlled by the state.53 Vietnamese media are state-owned, and regulated by the Ministry of Culture and Information (MCI).54 The MCI meets with journalists to instruct them on covering stories to ensure that the media follow the dictates of the state.55 In telecommunications, the Ministry of Posts and Telematics primarily sets policy and regulates.56

Vietnam is reportedly tightening its press controls. Deutsche Press Agentur states that a new Decree on Cultural and Information Activities, set to take effect July 1, 2006, punishes "denying revolutionary achievements," publishing stories that employ anonymous sources, disseminating reactionary ideology, refusing to permit someone interviewed for a story to read it before publication, or revealing state secrets of any type.57 The violations can be punished with fines of up to 30 million dong (roughly $2,000 U.S.).58 In addition, the new decree reportedly mandates that journalists submit articles for review before they can be published.59 The new, stricter press controls are apparently a state response to investigative journalism that has probed health issues, poverty, and governmental corruption, including an embezzlement scandal that led to the resignation of the minister of transportation and the arrest of the deputy minister.60

2. Internet Access Regulation

Legal regulation of the Internet in Vietnam occurs at several levels of government.61 Most broadly, Vietnam's National Assembly has adopted a number of decisions and decrees on information technology issues.62 Next, several ministries have authority to promulgate decisions affecting the Internet, including the Ministry of Finance,63 the Ministry of Post and Telematics (MPT),64 and the Ministry of Culture and Information (MCI).65 These ministries often collaborate on Internet regulation, including a circular regarding management of Internet agencies (such as cybercafés) promulgated in July 2005 that involved the MPT, MCI, Ministry of Public Security, and Ministry of Planning and Investment.66 In addition, provinces and cities regulate Internet services, particularly when they perceive a void in national regulation.67 Finally, the Vietnam Posts and Telecommunications Group (VNPT) makes decisions through its management that govern day-to-day functioning of the Internet infrastructure.68 Thus, control over Internet access and content can occur through national policymaking, ministerial decisions, or through rules created by the management of the backbone provider.

At the formal level, the state divides regulatory responsibilities among a number of ministries:

Ministry Regulatory Competence
Ministry of Posts and Telematics (MPT) Regulates state management of the Internet; coordinates with ministries, bodies at the ministerial level, state agencies, and People's Committees of centrally-run cities and provinces; represents Vietnam in international Internet activities;69 creates Internet development policies and planning; manages dialing and connection services; works with the state's Encryption Department to manage authentication. 70
Ministry of Science, Technology, and Environment Manages research and development of advanced Internet technologies.71
Ministry of Culture and Information (MCI) Manages information on the Internet, including implementing regulations for newspaper, publication, and e-news distribution online.72
Ministry of Public Security Ensures national security related to Internet activity and applies "technical measures" to manage information security online.73
Ministry of Finance Works with related branches of the state to create a financial support system for reduced Internet access charges for entities such as educational, health care, and software development organizations.74
Ministry of Planning and Investment Oversees granting and withdrawal of certificates for entities such as Internet agency businesses.75
Various ministries, ministerial-level bodies, and state agencies Are responsible for Internet application services in their area of regulation, including defining which services are forbidden.76
People's Committees Manage the implementation of Internet regulations in their cities and provinces in coordination with other state entities.77

Figure 5 ? Vietnam's Ministries and Their Regulatory Areas

Vietnam's most recent Internet access regulation derives from a joint circular issued by the MPT, MCI, Ministry of Public Security, and Ministry of Planning on July 14, 2005 that addresses both on-line access and content.78 The circular applies to Internet agents (Vietnamese individuals and organizations providing Internet access or on-line services), ISPs, and telecommunications online service providers.79 It bans using circumvention tools to bypass filtering, aiding others to do so, and accessing foreign ISPs via direct dial-up.80 Entities offering Internet access or on-line services to users must hold Internet agency registration certificates and must have entered into an agency contract with an ISP.81 The circular sets forth conditions that Internet agency businesses must meet to operate, such as ensuring that computer monitors can be monitored easily by managers.82

The circular also sets forth measures to control users' activities while on-line. Internet agency businesses must install centralized server software to ensure "information security" ? to filter undesirable Web sites.83 Businesses must post prohibitions and regulations on Internet use in a conspicuous place.84 Employees and owners of these businesses are responsible for inspecting users' activities and guiding them to comply with applicable regulations, and to prevent, detect, and stop any violations.85 Internet agency enterprises must compile information on their customers, including full names, addresses, identity card or passport numbers, the computer used by each customer, and the length of time they used the PC.86 Such information must be preserved, and shared with state agencies upon request.87
These businesses must also use technical methods to control Internet users' activities. They must install and use software to prevent users from accessing Web sites with prohibited material.88 The firms are required to employ software that records which sites users visit, what applications they use (for example, e-mail or chat), and how long they stay on-line.89 The circular requires businesses to store this information for 30 days, and contemplates that they will transmit it to state agencies responsible for information security.90

The circular also imposes obligations upon customers who access the Internet through these businesses. For example, they must not use software that allows them to illegally access or use Internet services (such as circumvention tools).91

ISPs and Internet agents can be inspected by the relevant state agencies; if they violate applicable regulations, they face penalties ranging from fines to termination of service contracts and licenses to potential criminal liability.92

Vietnam's legal regulation of Internet access is thus multi-layered and complex; while these rules may not be consistently implemented and enforced, they provide the state with considerable authority and discretion to control how citizens get on-line.

3. Internet Content Regulation

Vietnam seeks to ensure that Internet use supports the state's economic development while not threatening its one-party political system. The MPT reiterated these goals in a set of regulations promulgated in August 2005, which emphasized that it is unlawful to use Internet resources to oppose the state; to destabilize Vietnam's security, economy, or social order; to infringe organizations' or individuals' rights; or to interfere with the state's Domain Name System (DNS) servers.93 In addition, the decision makes clear that Internet operations supporting the state, such as those of the Communist Party or government agencies, have priority in resource allocation.94

Vietnam's approach to Internet content control is expressly multi-modal; the Deputy Minister of Culture and Industry stated that the state "should manage content in three ways: laws, techniques and education."95 The state has extensive legal regulation of Internet content and of the entities that provide it.96 Content providers in Vietnam must not host material that "cause[s] harms to the independence, sovereignty and territorial integrity of the Socialist Republic of Vietnam . . . [or] incite[s] people to oppose the State of the Socialist Republic of Vietnam or to destroy the entire-population great unity bloc."97 (Among other entities, this regulation applies to on-line newspapers.98) Likewise, they may not host content that "instigate[s] violence, propagate[s] aggressive wars, provoke[s] hostilities among nationalities and peoples of different countries, nor incite[s] obscenity, depraved life or crimes," or "disclose[s] the State's secrets, military, security, economic and diplomatic secrets as well as other secrets prescribed by the legislation of the Socialist Republic of Vietnam."99

The major guiding legal instrument regulating the Internet in Vietnam is the state's decree of Aug. 23, 2001 on Internet management, provision, and use.100 It covers all domestic and foreign individuals and organizations involved in Internet activity in Vietnam.101 The decree seeks to balance maintaining security and cultural values with developing scientific, research, educational, health-related, governmental, and software development uses of the Internet.102

In the decree, the state mandated that information stored on, sent over, or retrieved from the Internet comply with Vietnam's Press Law, Publication Law, Ordinance Protecting State Secrets, and other laws, such as intellectual property protections.103 Entities are responsible legally not only for information they create or send, but also information they store ? placing liability on intermediaries such as ISPs as well as authors.104 The state regulates the use of encryption on the Internet.105 Article 11 of the decree forbids damaging computer equipment services to prevent access to and use of Internet services; stealing and using passwords, codes, cryptographic keys, and private information from individuals or organizations; using the Internet "to fight against the Socialist Republic of Vietnam"; "causing chaos and security disorder"; "degrading national virtues and traditional good customs"; and "committing other violations of the laws."106 Furthermore, Vietnamese users may not use international direct telephone dial connections to access the Internet.107
The decree sets forth a schedule of fines for violations of Internet rules, including accessing the Internet unlawfully and using the Internet to spread pornographic images.108 Perhaps most important, users "who take advantage of the Internet to fight against the Socialist Republic of Vietnam, cause chaos or security disorder[, or commit] other serious violations that show signs of crimes will be subject to criminal penalties in accordance with the law."109

Under the 2001 decree, ICPs must obtain licenses from the Ministry of Culture and Information (MCI), and must follow the decree, MCI regulations on distribution of newspapers and publications online, and regulations on providing e-news on the Internet.110 ISPs and IXPs must obtain licenses from the MPT, and OSPs are regulated by the state agency responsible for that OSP's field.111 MPT regulations that implement the 2001 decree permit private ISPs to "provide accessibility to all Internet services at the request of members, except for service forbidden or not yet allowed to be provided, as well as Internet addresses forbidden to access as regulated and announced by State management agencies."112 This regulation forbids private ISPs from permitting access to filtered material.

One interesting result of the state's organization under the August 2001 decree is that regulatory responsibility for Internet material is divided along subject matter lines in Vietnam. While the Ministry of Culture and Information focuses on sexually explicit, superstitious, or violent content, the Ministry of Public Security monitors customers who access politically sensitive sites.113
Vietnam has recently moved to tighten its regulatory controls over the Internet. The state moved to clamp down on entities providing Internet information. Under regulations promulgated in 2002, domestic Web content providers must obtain approval and a license from the Ministry of Culture and Information.114 The Ministry of Public Security established a new police force "to monitor and suppress web sites carrying materials considered as crimes on the Internet" in August 2004.115 The new unit focuses on tracing sites that violate Vietnamese Internet laws.116

Vietnam has also sought to address security concerns related to Internet resources. In August 2005, the Minister of Posts and Telematics issued a decision on "The Management and Use of Internet Resources."117 This decision, which replaces a similar decision promulgated in 2003, regulates Internet resources such as IP addresses and domain names, and cautions that

It is strictly prohibited to use Internet resources for purposes of opposing the State of the Socialist Republic of Vietnam, disturbing security, economy, social order and safety, undermining the nation's fine tradition and custom, infringing upon the legitimate rights and interests of organizations and individuals and illegally obstructing the operations of the national system of domain name system (DNS) servers.118

Recently, the state moved to increase dramatically its supervision of cybercafés and other access points. In May 2004, Vietnam ordered its organizations to "tighten state management to prevent the exploitation and the circulation of bad and poisonous information on the Internet," instructing local authorities to inspect cybercafés and Web sites frequently. 119 For example, authorities reprimanded Hanoi cybercafé owners who allowed users to view forbidden Web sites, such as those containing pornography, after an investigation by the capitol's Culture and Information Department.120

The Ministry of Posts and Telematics issued a directive on May 7, 2004 "on intensifying the assurance of safety and security for post, telecommunication and Internet information in the new situation."121 This directive responded to the concern that "the hostile forces are increasingly taking advantage of postal, telecommunication and Internet services, to infringe upon our national security and undermine our social order and safety."122 It directed the MPT to implement information security and required "Postal, telecommunication and internet enterprises [to] coordinate with the local police in drawing up action plans to cope with violent incidents and riots undermining social safety and order or infringing upon national security."123 In addition, these entities were required to coordinate efforts with the Ministry of Public Security to block or withdraw services from entities using them to provoke "violence or riots."124 The directive requires that:

Postal, telecommunication and internet enterprises shall intensify examination and close supervision of activities of internet and telecommunication service agents; propagate and guide the agents' owners to firmly grasp the State's regulations on assurance of information safety and security and to strictly observe the regulations on provision of postal, telecommunication and internet services; [and] to supply fully data as requested so that the police agencies take necessary measures to protect national security and social safety and order.125

In July 2004, the Ministry issued a directive increasing control over Internet agents.126 The term "Internet agent" is defined in the August 2001 decree as the "Vietnam-based organizations and individuals who, on behalf of ISPs and PSPs, provide Internet dialing and Internet application services to users via agent contracts to enjoy[ing] commissions."127 While lauding the increased access provided by "public Internet agents" (such as cybercafés), the MPT decried those that allowed customers "arbitrarily [to] access unhealthy information which may cause harms to the political security, social order and safety and national cultural identity of Vietnam."128 Thus, the Ministry required owners of such agents to post the regulations on Internet services in their establishments, guide users in complying with the regulations, and stop known violations by users of the rules.129 ISPs and telecommunications providers were expressly permitted to suspend service to public Internet agents who facilitated or intentionally hid users' actions in "accessing electronic news pages or propagating or spreading documents with depraved or obscene contents, [or] violating the fine customs and traditions or infringing upon the national security."130 Furthermore, providers became responsible for inspecting and supervising agents in carrying out the terms of their contracts. If the provider detected a cybercafé violating contractual provisions, the ISP would have to stop providing service to the café and to inform the local Post and Telematics Department in writing.131 Providers with insufficient management and oversight of agents would face suspension of their ability to enter into new agent contracts.132

Vietnam augmented this directive one year later with a joint circular promulgated by the MPT, MCI, Ministry of Public Security, and the Ministry of Planning and Investment.133 Many of the circular's provisions recapitulate or expand upon items set forth in the 2004 directive. The circular prohibits storing military, economic, security, or other secrets on Internet-connected computers, using the Internet to oppose the Vietnamese state or disturb security, or creating Web sites or on-line forums to enable others to perform such prohibited activities.134 Users are responsible for content they create on or transmit through the Internet.135 They must not transmit depraved or obscene information or images, "offend fine traditions and custom," oppose the Vietnamese state disturb social order, or use software to access and use Internet services unlawfully through their Internet activities.136 Moreover, the regulations deputize users as detectives in discovering prohibited material on-line; if a user finds a Web site or service with content that opposes the Vietnamese state, or disturbs social order and security, she is responsible for informing promptly the municipal or provincial Culture and Information Services (a department of the MCI).137

Recently, Vietnamese authorities further increased monitoring of cybercafé usage. The Ministry of Culture and Information promulgated an instruction in August 2004 requiring cybercafés to track all Web sites their customers visit, along with the credit card or ID card numbers of customers who visit inappropriate sites, and to make this information available to its inspectors.138 However, ONI research inside Vietnam demonstrates that enforcement of ID regulations varies greatly in different locations and is, for example, largely ignored in cities such as Ho Chi Minh City and Hanoi. During ONI's in-country research, there were no instances of cybercafé users being asked or required to present identification, and the ONI researcher did not observe any use of surveillance cameras or equipment to monitor Internet usage. While most cafés posted the Internet café regulations (N?i Qui Ph?ng M?y) on their walls, users did not appear to read or pay attention to them. The regulations comprise the following:139

Internet Service Regulations:140

  1. Do not act in an unruly manner, do not sabotage the equipment, and do not distort, alter or destroy the Internet programs on the computers.
  2. Do not create and purposely spread viruses on the Internet.
  3. Do not use the software to search and use illegal services provided on the Internet.
  4. Do not steal and use passwords or private documents belonging to organizations and individuals, or distribute them for use to others.
  5. Do not use the Internet to threaten, blackmail and damage the reputation of others.
  6. Do not use the Internet to distribute materials against the interests of the government or the national security of Vietnam.
  7. Do not use the Internet to post pornographic materials and pictures contrary to Vietnamese morals, and do not put online materials that are in violation of the publishing law.
  8. Do not contact foreign companies providing Internet services by using dial-up direct phone calls.
  9. Do not use any Internet programs currently prohibited or not yet authorized for use in Vietnam.

The size of partitions between computers varied from substantial to non-existent, and some cafés posted small signs (in English) on computer monitors asking users not to visit pornographic Web sites.141 Fines for café owners were increased to a maximum of 30 million dong (U.S. $1900).142 State officials claimed that their monitoring of cybercafés was difficult due to a lack of equipment and technology expertise.143 In September 2004, authorities shut down at least 65 cybercafés in Ho Chi Minh City because patrons visited pornographic and political opposition Web sites.144

In March 2005, the MCI began "inspecting" the approximately seventy state-run online newspapers and informational websites, and more than 2,500 other news and informational websites, under the MCI's control.145 The goal of the inspection was to locate those news and information websites that "operate outside the rules."146 The Deputy Minister of the MCI admitted to the state-run Internet news service VietnamNet that no single agency can effectively police all the online news and information websites in Vietnam.147 He suggested that more regulations are needed to "govern the control and provision of information on the Internet," including "regulations that clearly define the conception of an online newspaper, information websites and ICPs."148 In addition, he called for the law to provide broader power to impose sanctions on a range of online activities: "While we focus on violations of public Internet agents, a series of new violations have appeared, for example chat sex, sex films on mobile phones, etc."149

Thus, the trend in Vietnam is towards increased regulation of Internet content, resources, and access. Vietnamese authorities have launched a crackdown on what they consider unlawful usage of the Internet in an effort to prevent citizens from accessing material that might undermine its control. Backed up by explicit regulations, these controls are distributed across multiple government ministries and extend into the operations of major ISPs, cybercafés, and Web site operators.

3. TESTING METHODOLOGY


A. Methods

ONI performs technical testing across multiple levels of access at multiple time intervals in a number of regions around the world. The team analyzes results within the contextual framework of the target state's filtering technology, law, and regulations. To obtain meaningful, accurate results we seek to:

  • generate lists of domain names and URLs that have been or are likely to be blocked, based upon background research into relevant social and political issues in Vietnam;
  • enumerate ISPs and national routing topography;
  • determine the type, location, and behavior of the filtering technology;
  • deploy network interrogation and enumeration software at multiple access points; and,
  • conduct a thorough statistical analysis of results.

Determining which URLs to test is a vital component of our research, as it reveals the filtering system's technical capacity and content areas subject to blocking. ONI employs three types of lists:

  • a list of "high impact" sites reported to be blocked or likely to be blocked in the state due to their content (for example, political opposition);
  • a "global list" containing a control list of manually categorized Web sites reflecting a range of Internet content (for example, news and hacking sites), intended in part to enable comparisons across multiple states; and
  • a multilingual list of significant key words used to generate significantly larger lists through search engine queries.

To explore Internet filtering, we deploy network interrogation devices and applications, which perform the censorship enumeration, at various Internet access levels. These tools download the ONI testing lists and check whether specific URLs and domains were accessible from that point on the network. Interrogation devices are designed to run inside a state (i.e., behind its firewall) to perform specific, sensitive functions with varying degrees of stealth. Similarly, ONI distributed interrogation applications to trusted volunteers who ran the software inside the state. For testing, depending upon a series of local factors, ONI obtains network access at multiple levels through a combination of:

  • Proxy servers,
  • Long distance dial-up,
  • Distributed applications, and
  • Dedicated servers.

During initial testing, we use remote computers located in states that filter. These remote computers are located behind the state's firewalls yet allowed access to clients connecting from the wider Internet. We attempt to access the URL and domain name lists through these computers to reveal what content was filtered, and how consistently it was blocked. The ONI team also tests these lists from control locations in non-filtered countries. The testing system flags all URLs and domains that were accessible from the control location, but inaccessible from ones inside the target state, as potentially blocked.


B. Results Analysis

We carefully analyze the data obtained from testing to document the nature of filtered content, to explore the technical capabilities of the target state, and to determine areas that require in-depth study during testing from inside the state's firewall. In particular, ONI examines the response received over HTTP when attempting to access filtered content. As discussed, when content is filtered, users often receive a block page. In other cases, filtering can be less obvious or transparent, appearing to be network errors, redirections, or lengthy timeouts rather than deliberate blocking. We analyze HTTP headers ? text sent from the Web server to the browser ? to derive information about both the server and the requested page. This information is generally hidden from the end user. However, these headers can indicate whether content was successfully accessed or was inaccessible. If an error occurs, the HTTP protocol returns codes that indicate the type of error in the header. Thus, by analyzing the headers captured during testing, we seek to distinguish between errors caused by Internet filtering and more mundane, unintentional network connection errors.
We classify results in one of four categories:

  • URL is accessible both through the local connection and the remote computer (not filtered);
  • URL is accessible through the local connection but inaccessible through the remote computer, which returned a different HTTP response code (possibly filtered);
  • URL is accessible through the local connection but inaccessible through the remote computer due to a network connection error (possibly filtered, but not definitive); or
  • URL is accessible through the local connection but inaccessible through the remote computer; the remote computer returns a block page (filtered).

If a URL is inaccessible through both the local connection and the remote computer, we consider it "dead" and remove it from the results. This result indicates that the URL's content was not available to Internet users generally at the time of our testing, making the URL irrelevant for our purposes.

The ONI team analyzes blocked, unblocked, and uncertain URLs both at an aggregate level (to estimate the overall level of filtering) and at a category level (to indicate what types of content the state seeks to control). We publish state-specific studies, such as this one on Vietnam, that provide background on a state's political and legal system, lists of tested sites, and analysis of results to reveal and analyze, to the greatest extent we can given the data we are able to collect, what information a state blocks and how it does so. We note, however, that our results and analysis capture a "snapshot" of a state's filtering system for a specific point or period of time; governments can and do alter the content they block dynamically.

C. Methods Specific to Vietnam

In Vietnam, we tested from various access points (including hotel, cybercafé, and wireless connections) on two ISPs, FPT and VNPT, with the assistance of an anonymous in-country tester and an ONI researcher who traveled to Vietnam.150 The initial results, from the in-country volunteer, were obtained from FPT in November 2005. The more comprehensive testing, by the ONI researcher, was performed in March 2006 on both FPT and VNPT.

Results varied between the two networks. This network-dependent filtering may represent a centrally-distributed blacklist to ISPs, and some contacts with in-state experts suggest as much, though we are unable to say this with certainty. We believe, though, that filtering occurs at multiple levels in Vietnam, but not with a high degree of consistency. We also believe that Vietnam filters for individual and home users, but not for businesses. Since most users access the Internet through the FPT and VNPT networks, either directly (57% of ISP subscribers in Vietnam use VNPT, and 20% use FPT) or indirectly (FPT and VNPT are two of only six entities licensed to provide gateway access to the global Internet as IXPs), we believe our results are representative of the average user's experience.


D. Topics Tested

Topics to which the Vietnamese state has demonstrated sensitivity include democracy, political opposition to the current government, human rights, protests by the Montagnards (an ethnic group of primarily Christians who supported the U.S. during the Vietnam-U.S. war) in February 2001 and subsequently,151 the Sino-Vietnamese border accords,152 Buddhism,153 illegal Internet telephony,154 the proposed U.S. Vietnam Human Rights Act,155 and certain religious movements, including Dao Cao Dai (Caodaism).156 In addition, Vietnam has arrested a number of dissidents, whose cases have attracted international attention. These include Tran Khue,157 Nguyen Vu Viet, Nguyen Truc Cuong, Nguyen Thi Hoa, Nguyen Dan Que, Pham Que Duong, Nguyen Khac Toan, Pham Hong Son, Nguyen Vu Binh,158 Le Chi Quang, Pham Hong Quang, and Nguyen Dinh Huy.159 We tested Web sites related to these topics to discover whether Vietnam blocked access to them.

4. RESULTS AND ANALYSIS


A. Summary

Our testing of Vietnam's Internet filtering found that the state concentrates its blocking on political and religious topics. Surprisingly, Vietnam does not block any pornographic content (though it does filter one site with links to adult material), despite the state's putative focus on preventing access to sexually explicit material. The state's filtering concentrates on Vietnamese-language content, and often blocks several times as many sites on a given topic in Vietnamese as it does those in English. This is characteristic of states like Vietnam that create their own list of banned sites rather than relying upon a commercial software provider for block lists. Vietnam's ISPs implement filtering differently, with FPT relying on removing listings for prohibited sites from its Domain Name System (DNS) servers while VNPT uses a system that displays a "block page" when users try to access banned sites. FPT's mechanism is more subtle, and possibly cheaper to implement, but also potentially more susceptible to circumvention. Overall, FPT's blocking was greater than that on VNPT. Finally, FPT's blocking has increased over time ? both within content categories and in the number of such categories with filtered sites ? and ONI believes this trend will continue.

B. Global List Results

In order to enable comparison across states ONI studies, we have developed a global list that includes 869 sites in 29 categories.160 This list includes a variety of sites in content areas that are often targeted for filtering, including anonymizers (sites that enable users to circumvent filtering or to access material anonymously), pornography, religion, and human rights. We test this list in each state ONI studies. In the years ONI has studied filtering, we have evaluated changes in state-based blocking and in the sites on our global list. This research led ONI to update the global list, creating a second, revised global list. For backwards compatibility, we tested both the original and revised global lists in states such as Yemen. In this study, we have migrated to using solely the revised global list. While this limits ONI's ability to compare filtering to countries that we have previously studied (where we tested only our original global list), this change ensures that our testing is up-to-date and reliable. ONI will continue to employ the revised global list in our future studies.

In Vietnam, ONI's global list testing found little filtering on either network in March 2006. FPT blocked 12 sites out of 851 tested (1.4%), and VNPT filtered 5 sites of 821 tested (0.6%).161 Of the limited sites blocked, both FPT and VNPT concentrated their filtering on anonymizers ? sites that allow users to bypass filtering systems and to access prohibited content. FPT blocked 7 of 29 anonymizer sites tested (24%), and VNPT blocked 3 of 28 sites (11%).162 The anonymizer sites blocked by VNPT were also blocked on FPT.

Our results found sporadic, minimal blocking in other content categories. FPT filtered two Free Web Space sites (of 30 tested, 7%); one Human Rights site (of 30 tested, 3%), that of Human Rights Watch; one Religion site (of 30, 3%), that of Catholic Online; and one site related to an entity on the U.S. State Department's list of terrorist organizations (of 44, 2%), the Communist Party of Peru (also known as the Shining Path). VNPT blocked one Free Web Space site (of 28 tested, 4%), though this site was available on FPT, and the same Human Rights site filtered on FPT (of 26, 4%; the site is that of Human Rights Watch). The low level of blocking in other categories probably results from the state's focus on filtering Vietnamese-language, rather than English-language, content.

Notably, though, neither ISP blocked any of the pornographic sites tested, nor any of the other sites with sex-related content (such as sex education, provocative attire, gay / lesbian / bisexual, or dating sites), despite the state's putative focus on pornography as a justification for its filtering efforts. Informal discussions with knowledgeable sources in Vietnam indicate that this is not an artifact of language (most of the pornographic sites tested are in English), as similar sites that feature sexually explicit material in Vietnamese are not filtered either.

Overall, FPT's blocking in our global list testing was more extensive than VNPT's blocking. In every category that had one or more blocked sites, except one (Human Rights), FPT blocked more sites than VNPT did; in the Human Rights category, both ISPs blocked a single site. The global list was thus a microcosm of ONI's overall findings, which show that FPT's blocking is more extensive than VNPT's, both in the types of content filtered and in the extent of sites blocked within a given category.

ONI initially tested filtering on FPT using an in-state volunteer in November 2005. Those tests found no filtering on FPT at that date; thus, even the relatively low level of blocking by the ISP represents a considerable increase in content control in a relatively short period of time (roughly six months). This pattern fits ONI's conclusion about filtering in Vietnam generally: it is increasing, becoming more effective and more widespread.


C. Vietnam-Specific Results

To investigate Internet filtering on topics sensitive to the Vietnamese state, we tested three state-specific lists in addition to the global list: a high-impact list of sites known or alleged to be filtered, a list derived from links on the blocked Saigon Bao Web site, and a long list of sites derived from the top results of Google queries for sensitive terms (such as the name of Vietnamese dissidents). While we attempted to test all these Vietnam-specific lists on both ISPs, access constraints prevented us from obtaining complete results for the long list on VNPT. Thus, the analysis for the long list below concentrates on FPT, while noting where our limited results allow us to draw conclusions about VNPT's filtering. Analysis for the high-impact and Saigon Bao lists covers both ISPs.

1. High-Impact List

ONI developed a list of high-impact Web sites that focus on material on issues sensitive to the Vietnamese state. Many of these sites are known or alleged to have been blocked in the past. The list contains 85 sites in 18 categories. ONI tested the high-impact list in November 2005 on FPT with the aid of an in-state volunteer, and an ONI researcher tested the list on both FPT and VNPT in March 2006. Thus, we can compare the filtering between ISPs as of March 2006, and can also perform longitudinal analysis to elucidate changes to FPT's filtering practices in the six months between our rounds of testing.
a. Comparison Between FPT and VNPT in March 2006
Overall, FPT blocked 47% of accessible sites on the high-impact list in March 2006 (39 of 83 sites, with 4 not accessible for unrelated reasons), while VNPT blocked 39.8% of accessible sites (33 of 83, with 4 not accessible). In every category where at least one site was blocked, FPT's filtering was equal to or greater than VNPT's filtering. FPT's blocking was noticeably higher in the politically-oriented categories of Dissidents and Human Rights. Both ISPs blocked nearly all Pro-Democracy and Political Opposition sites, along with a significant fraction of News and Dissident sites.

Figure 6
Figure 6
High-Impact List Filtering by ISP

Our high-impact results also found limited filtering of Buddhist and Caodai religious material, along with sites on the Montagnard people and overseas Vietnamese communities. ONI's testing returned blocks of the single site we tested in the Forum, Freedom of Expression / Journalism, and Erotic categories; the filtered erotic site was the only blocking of sexually-themed material we found during our testing, despite Vietnam's purported focus on that type of content for its filtering regime.

High-Impact Category FPT Sites Blocked / Tested (% Blocked) VNPT Sites Blocked / Tested (% Blocked)
Erotic 1/1 (100%) 1/1 (100%)
Forum 1/1 (100%) 1/1 (100%)
Pro-Democracy 9/9 (100%) 8/9 (88.9%)
Political Opposition 7/9 (77.8%) 7/9 (77.8%)
Human Rights 8/11 (72.7%) 6/11 (54.6%)
Dissidents 6/12 (50%) 4/12 (33.3%)
Freedom of Expression / Journalism 1/2 (50%) 1/2 (50%)
Montagnard 1/2 (50%) 1/2 (50%)
News 5/10 (50%) 3/10 (30%)
Overseas Vietnamese 1/2 (50%) 0/2 (0%)
Religion - Buddhist 2/6 (33.3%) 0/6 (0%)
Religion - Caodai 1/5 (20%) 1/5 (20%)
Religion - Christian 0/7 (0%) 0/7 (0%)
Religion - Muslim 0/1 (0%) 0/1 (0%)
Social Issues 0/2 (0%) 0/2 (0%)

Figure 7
High-Impact List Filtering by Category

b. Longitudinal Analysis of Changes in FPT Filtering
ONI tested FPT's filtering from within Vietnam using our high-impact list in early November 2005 and late March 2006. Though this is a difference of only roughly six months, we found a noticeable increase in filtering on the FPT network during that time. ONI found an increase in overall measures of filtering, such as the number of content categories in which least one site was filtered and the aggregate level of filtering.

Criterion November 2005 March 2006
Overall filtering (sites blocked / tested, % blocked) 15 / 78 (19.2%) 39 / 83 (47%)
Content categories with at least one site blocked 5 (of 18) 11 (of 18)

Figure 8
Filtering Metrics for FPT Over Time

FPT's filtering also increased in every content category with at least one filtered site (except one, Montagnards, which remained unchanged).

Category November 2005
(sites blocked / tested, % blocked)
March 2006
(sites blocked / tested, % blocked)
Pro-Democracy 1/9 (11.1%) 9/9 (100%)
Dissidents 0/11 (0%) 6/12 (50%)
Political Opposition 2/7 (28.6%) 7/9 (77.8%)
Religion ? Buddhist 0/6 (0%) 2/6 (33.3%)
News 3/9 (33.3%) 5/10 (50%)
Freedom of Expression / Journalism 0/1 (0%) 1/2 (50%)
Overseas Vietnamese 0/2 (0%) 1/2 (50%)
Human Rights 7/11 (63.6%) 8/11 (72.7%)
Forum 0/1 (0%) 1/1 (100%)
Erotic 0/1 (0%) 1/1 (100%)

Figure 9
Filtering by High-Impact Category for FPT Over Time

c. Language Effects
ONI's analysis of the results of our high-impact list testing discerned a pattern in Vietnam's blocking that held over time and across our different lists: the state filters Vietnamese-language sites to a greater degree than it does English-language ones, both overall and in various content areas. (Since some sites are available in more than one language, the chart below counts a site in each language category in which it offers material.) As described in detail below, we believe this pattern results from Vietnam's decision to create its own block lists rather than to rely upon a commercial software program for these lists. In selecting sites to filter, Vietnam unsurprisingly focuses upon sites that offer sensitive material in the state's most commonly-used language.

Figure 10
Figure 10
Filtering of High-Impact Sites by Language

d. Summary
FPT's filtering has increased over time, both in the breadth of content blocked and in the level of blocking for a given category of material. Given that the number, percentage, and number of categories of sites filtered by FPT more than doubled in the six-month period between our tests, ONI believes it is likely that FPT's level of blocking will continue to increase.

2. Saigon Bao List

We compiled a list of sites linked from the Web site www.saigonbao.com, which lists sensitive sites such as those of opposition media sources. (The Web site itself is blocked by both FPT and VNPT.) Our testing found high levels of blocking for sites in categories such as opposition politics, politics, and human rights and democracy.

Category FPT VNPT
Sites Tested / Blocked (% Blocked) Sites Tested / Blocked % Blocked
General News 1 / 23 (4.3%) 0/23 (0%)
Human Rights & Democracy 39 / 128 (30.8%) 31 / 126 (24.6%)
Opposition Politics 4 / 13 (30.8%) 3 / 12 (25%)
Politics 19 / 34 (55.9%) 16 / 33 (48.5%)
Portal Sites 0/1 (0%) 0/1 (0%)
Religion 2/2 (100%) 1/2 (50%)
Vietnam News 0/2 (0%) 0/2 (0%)
Total 65 / 203 (32%) 51 / 199 (25.6%)

Figure 11
Filtering of Selected Sensitive Sites from SaigonBao.com

While the selection of these URLs targeted sites believed to be blocked by Vietnamese ISPs, ONI believes that the pattern of blocking conforms to the overall shape of the state's filtering regime by targeting content that could be politically damaging or destabilizing to the current government. This accounts for the high levels of blocking of politically-oriented, human rights-focused, and democracy-oriented Web sites. Religious Web sites also may be a focus of filtering, but our sample size was too limited to draw definitive conclusions about this type of content.

Figure 12
Figure 12
Graph of Blocked SaigonBao Categories by ISP

The SaigonBao testing results also fit the pattern for the two ISPs found in our high-impact testing: FPT's filtering was more extensive than VNPT's for every category in which one or more sites was blocked. As mentioned above, ONI believes that the trend for Vietnam is towards more extensive blocking, and that future testing will likely reveal a shift in VNPT's filtering in the direction of more comprehensive filtering of material.

3. Long List

ONI compiled a long list of over 2000 URLs on sensitive topics to provide a more extensive picture of Vietnam's filtering regime. We created this list using Google queries for key terms (such as "L? Ch? Quang" Vi?t") and the Open Directory Project classification system. (The Open Directory Project, known as "dmoz," constructed a massive taxonomy of Internet content and uses volunteers to maintain lists of the most useful, content-rich sites in each category.163)

We were able to test the long list on the FPT network from within Vietnam in early November 2005 and late March 2006. Due to access constraints, we were unable to obtain a complete run of the long list on VNPT; however, we did obtain partial results, and the analysis below notes VNPT trends where there is sufficient data to support such an inference.

We combined our testing results with our high-impact list results, removing duplicate URLs, and then analyzed the data. Below, we compare results on FPT between the two testing runs to reveal changes in the ISP's over time, examine differences in blocking based on a Web site's language, look at certain inconsistencies in FPT's blocking, and also describe the content areas upon which the filtering focuses.

a. Longitudinal Analysis
ONI's long list testing, consistent with the results from our other testing lists, showed that filtering on FPT has increased over time. In November 2005, we checked 1988 URLs and found 282 blocked (14.2%). In March 2006, we checked 2012 URLs and found 599 blocked (29.8%).164 FPT's filtering had thus more than doubled in six months.

b. Language Effects
ONI obtained data on the language used by the Web sites we tested. The pattern was clear: Vietnamese-language content was blocked at a higher rate than English-language content.165 This trend held over time for FPT and was largely similar on VNPT, although our smaller sample size for VNPT limits our ability to generalize about that ISP. For example, blocking was often several times greater for Vietnamese-language sites than for English-language ones for key categories such as Politics, Dissidents, Religion, and Media.

Figure 13
Figure 13
FPT Filtering by Language (November 2005)

The pattern of greater filtering of sites in Vietnamese rather than in English continued during our March 2006 testing.

Figure 14
Figure 14
FPT Filtering by Language (March 2006)

The pattern on VNPT was similar, although the number of sites tested was far smaller. One category, Religion, showed greater filtering of English-language Web sites than Vietnamese-language, but this may be a result of the small sample size (3 sites tested in English, 6 sites tested in Vietnamese, 1 site blocked in each language).

Figure 15
Figure 15
VNPT Filtering by Language (March 2006)

Greater filtering in a state's primary language rather than in English-language sites is characteristic of states that create their own list of sites to block rather than employing commercial Web filtering software. Commercial software makes it easy for a filtering state to block most of certain content categories ? for example, pornography or illegal drugs material ? quickly and easily. The software tends to focus on English-language content, though, for three reasons: most major Web filtering software companies are located in English-language states; a large fraction of the Web's pages in categories that are frequently targeted for blocking are English-language; and commercial software providers rarely (if ever) produce lists tailored to a particular state's filtering goals (for example, Vietnamese political opposition) due to the limited market for such items.
States that create their own block lists naturally focus upon sites that are on sensitive topics and that are most accessible or likely to be viewed by their citizens. This leads to an emphasis on sites in the state's native language, as those are most likely to be relevant and interesting to its citizens.
ONI believes that Vietnam does not currently employ commercial filtering software. The pattern of blocking does not fit that of any software product that ONI has studied, and FPT's method (described below) of deleting DNS records for prohibited sites is inconsistent with using Web filtering software. We believe that the greater filtering of Vietnamese-language sites on a given topic (corresponding to a keyword search or dmoz category) compared to English-language sites derives from either the Vietnamese state or Vietnamese ISPs compiling their own block lists rather than relying on those obtained from commercial Web filtering software providers.

c. Filtering Inconsistencies
Another characteristic of Vietnam's filtering system that suggests its block lists are compiled by the state or ISPs, rather than by a commercial software firm, is that its blocking is occasionally inconsistent. A common trait of individually-created block lists is that they fail to filter all of the ways by which users can access a single site. For example, ONI's Web site can be accessed by entering www.opennet.net or opennet.net. Filtering the complete URL (www.opennet.net) but not the domain name (opennet.net) will prevent some, but not all, users from reaching ONI's site. Both of Vietnam's ISPs demonstrate this inconsistency, leading to partial filtering of some Web sites.

In our November 2005 testing, for example, FPT demonstrated this behavior with three Web sites. The sites www.lenduong.net, www.nguoi-viet.com, and www.vps.org were blocked at those URLs, but were accessible at the URLs lenduong.net, nguoi-viet.com, and vps.org. Similarly, the site quehuongmedia.com was blocked at that URL, but available at www.quehuongmedia.com. This inconsistency allowed knowledgeable Internet users to circumvent filtering by trying multiple forms of a URL ? an FPT user blocked from reaching www.lenduong.net could try, and succeed, in reaching the same site (that of the International Vietnamese Youth Network, which "promot[es] human rights and democracy") at the URL lenduong.net.166 In effect, this partial blocking allows a low-technology form of circumvention.

However, by the time of our March 2006 testing, FPT had plugged these holes: all four of the Web sites described above were blocked at both forms of their URLs. FPT's filtering had improved, and the easy circumvention made possible earlier by changing the URL entered in the browser no longer functioned.

VNPT continued to display at least one instance of this filtering inconsistency during our March 2006 testing. The site www.vps.org (the site of the Vietnamese Professionals Society, which advocates for democracy in Vietnam) was blocked at that URL, but available at vps.org.167 Note that ONI observed this same behavior for FPT for this site during our November 2005 testing, but found it was resolved by our March 2006 testing.

This type of learning is common to filtering systems that refine their block lists experientially - through trial and error - rather than through the comprehensive, automated techniques that commercial filtering companies employ to prevent users from easily bypassing their blocks. We believe that, like FPT, VNPT will refine and tighten its filtering system to address such workarounds.

d. Targeted Content
Overall, our long list data confirms the pattern observed in our high-impact and SaigonBao list testing: Vietnam's filtering efforts concentrate on politically and religiously sensitive material. (There is often an overlap in these categories in Vietnam; for example, the Montagnard people, who assisted the United States during the Vietnam War, are frequently adherents of Christianity.168) Political topics such as opposition to the current Communist government, human rights, and the Vietnam Human Rights Act are targeted in particular for blocking. Our FPT testing of the top sites corresponding to a query for the Vietnam Human Rights Act on Google, for example, found 19% of English-language sites blocked (17 of 89) and 63% of Vietnamese-language sites filtered (56 of 89).169

Figure 16
Figure 16
Political Content Blocking on FPT

The state also censors on-line content related to dissidents. As with other filtered subjects, ONI found that material is more likely to be blocked if it is in Vietnamese than in English. (Our testing results for VNPT had a sample size too small to draw conclusions regarding dissident sites.)

Figure 17
Figure 17
Blocking of Dissident Sites on FPT by Language

Religious materials - particularly those related to faiths with a strong presence in Vietnam, such as Buddhism and Caodai - are filtered to a lesser, but still significant, degree. Again, filtering of Vietnamese-language sites was greater than that of English-language sites. For example, while only 2% of the top English-language Google results for the term "Buddhism" were blocked (2 of 96 tested), 34% of the top Vietnamese-language Google results for the corresponding term "d?o Ph?t" were filtered (32 of 95). ONI observed a similar pattern for Google results related to Buddhism in Vietnam, with 8% of English-language sites filtered (8 of 97), but 35% of Vietnamese-language sites blocked (31 of 88, search term "Vi?t "d?o Ph?t""). Caodai sites followed this trend also: 6% of top English-language sites for a Google query (6 of 99), but 34% of Vietnamese-language results (32 of 93, search term "Cao ??i Vi?t "). Finally, sites obtained from a Google search related to the Montagnards were blocked 13% of the time for English-language results (13 of 98), but filtered 58% of the time for Vietnamese-language ones (38 of 65, search term "d?n t?c Vi?t Nam montagnard"). Overall, while religious materials are filtered less extensively than political content, Vietnam blocks a significant fraction of religious sites, with particular attention to those that host Vietnamese-language content.

Figure 18
Figure 18
Religious Content Blocking on FPT

While our analysis concentrates on FPT, the limited data available suggests that VNPT's filtering targets the same material, but that its filtering is often less pervasive than that on FPT.

Figure 19
Figure 19
Comparison of Political Blocking by ISP
Figure 20
Figure 20
Comparison of Religious Blocking by ISP

Finally, our longitudinal analysis of blocking on FPT from November 2005 to March 2006 confirms the pattern described in the sections above: FPT's filtering has become more extensive, often considerably so.

Figure 21
Figure 21
Changes in Blocking on FPT Over Time

This trend of increased filtering also applied to specific content categories. For example, material on Vietnamese dissidents became more difficult to access from November 2005 to March 2006 on FPT; in many cases, the blocking roughly doubled.

Figure 22
Figure 22
Changes in FPT Filtering of Dissident Sites Over Time

Similarly, the filtering of political topics tightened during this six-month period. For example, the blocking of sites related to the Vietnam Human Rights Act increased for both English-language and Vietnamese-language URLs.

Figure 23
Figure 23
Comparison of Filtering of VHRA Sites Over Time on FPT

ONI observed the same increase for sites with religious material, regardless of language.

Figure 24
Figure 24
Comparison of Filtering of Religious Sites Over Time on FPT

For FPT, the overall conclusion is clear: filtering has increased with time. ONI believes this trend will continue.


D. Filtering Methods

The ISPs that ONI tested, FPT and VNPT, utilize different filtering techniques. When we attempted to access a filtered site from VNPT, the system returned a "block page" indicating that the requested site was prohibited.170 Thus, while users are prevented from accessing filtered material, they are at least informed that this is a deliberate step and not the result of a technical failure or error.

By contrast, when we attempted to access a filtered site from FPT, the browser indicated that the requested site did not exist. When we queried FPT's DNS (Domain Name System) server, we found that the listings (A records mapping a site's URL to its IP address) for filtered sites had been removed.171 A request for a blocked site, such as ykien.net, would therefore result in this error message because the browser was unable to locate the server for that Web site. This form of filtering is more subtle and insidious than the block page method because it disguises a deliberate attempt to prevent access to certain content as a technical error or as a failure by the user in requesting a non-existent site. In addition, it is likely cheaper to implement effectively than a commercial software system that uses a block page. However, this method is also potentially more vulnerable to circumvention by technically knowledgeable users than VNPT's mechanism is. A user on FPT could modify her computer to use a DNS server on a different network than FPT; her DNS requests for prohibited sites would return an IP address that would allow her browser to locate and connect to those sites. While this circumvention technique requires understanding both how FPT blocks access to sites, and how to modify which DNS server one's computer uses, it is relatively straightforward to employ.

The two ISPs thus use different mechanisms to effect filtering. VNPT employs filtering software that, upon receiving a request for a blocked site, returns a block page to the requesting user. FPT removes records for filtered sites from its DNS, so users who attempt to access these sites receive an error message ? using filtering software is unnecessary.


5. CONCLUSION

Despite its putative focus on sexually-explicit content, Vietnam's Internet filtering regime blocks virtually none of this material, focusing instead on politically and religiously sensitive sites that could challenge the state's one-party system. Web sites that contain information on topics such as Vietnamese political dissidents, human rights, political opposition to the current government, the proposed Vietnam Human Rights Act in the United States Congress, and human rights are subject to significant filtering that renders them inaccessible for most users in Vietnam. Religious content, such as pages on religious freedom, Buddhism, and Caodai are blocked to a lesser, but still significant, degree. Some topics, such as the Montagnard people who assisted the United States during the war with Vietnam and who are frequently Christians, overlap these categories and are filtered accordingly. Over time, the state's on-line filtering has expanded, both in the content blocked for a given topic and the number of content categories that are targeted. However, Vietnam did not block any of the pornographic or other sex-related Web sites ONI checked, with the exception of a single site with links to erotic materials; the state's filtering practices are thus in obvious tension with the purported justification for these actions. The filtering regime thus fails to provide even a figurative fig leaf to cover the state's rationale for blocking.

Unlike many states that implement Internet filtering, Vietnam does not employ commercial filtering software, relying instead upon its own block lists of prohibited sites. It implements this filtering differently on different ISPs. ONI found that FPT, for example, blocks more content, in more topic areas, than VNPT does. Given the increase in FPT's filtering over time, though, ONI believes that VNPT's filtering will move towards FPT's pattern by expanding.

Moreover, the two ISPs use different methods to prevent access to banned sites. VNPT's system checks a user's request for a page against the block list; if the requested site is prohibited, the ISP returns a block page informing the user why the URL cannot be accessed. In contrast, FPT's filtering is more subtle ? it masquerades as an inability to locate the requested site. FPT removes entries for filtered sites from its DNS servers; users who try to reach these URLs receive a message indicating that the site does not exist. While this type of filtering masks the state's blocking efforts, and may be less expensive to implement, it is also potentially susceptible to circumvention by technically savvy users who can switch their computer settings to another DNS server. This is the first instance of DNS-based filtering that ONI has observed.

Vietnam's filtering efforts are plainly becoming more sophisticated and effective. Both ISPs tested make an effort to block some popular anonymizer services that permit users to bypass filtering. In addition, FPT has over time improved its consistency in blocking, weeding out instances where a site is filtered at one URL but available at another. This refinement, combined with the overall increase in the breadth and depth of the state's filtering, suggests that Vietnam is growing more successful in controlling what content its citizens can view and exchange. In short, Vietnam is moving up the learning curve in its design and implementation of Internet filtering.

The state focuses its efforts on Vietnamese-language content, while blocking of English language sites is less common. This concentration on local language material is characteristic of filtering regimes that develop their own lists of banned sites rather than outsourcing these decisions to commercial software providers. While commercial software makes it easy to block broad categories of material (such as pornography), locally-developed block lists are often more effective in preventing access to sensitive content in the language most likely to be used and understood by citizens.

Overall, Vietnam's filtering system accords with the state's larger focus on maintaining political control by blocking opposition while simultaneously pursuing economic growth. While the Vietnamese state's blocking of access to certain content on the Internet can be circumvented by users with technical knowledge, ordinary users will likely continue to find that filtering distorts their information environment, and even technically sophisticated Vietnamese will face legal penalties for accessing or distributing prohibited content. Vietnam's filtering regime is multi-layered, relying not only on computing technology, but also on threats of legal liability, state-based and private monitoring of users' on-line activities, and informal pressures such as supervision by employees or other users in a cybercafé. Vietnam is likely willing to accept that a small number of skilled users can bypass its technical filtering measures since its multi-modal approach will keep the vast majority of users within bounds of permissible behavior.

While Vietnam has fewer resources to devote to on-line content control than states such as China, the country has nonetheless established an effective and increasingly sophisticated Internet filtering system. Despite the stated rationale of targeting sexually explicit content, Vietnam ignores this material in its blocking, concentrating instead on political opposition and other topics that could undermine the current socialist system. The state's multi-pronged approach to information control, improved effectiveness in filtering sites, and continued concentration on the Internet as a key communications medium suggest that Vietnam will continue its move towards increasingly comprehensive filtering.


Appendix 1

Vietnam High-Impact List Testing Results (March 2006)

Appendix 1

Languages and Descriptions of URLs on High-Impact List

Appendix 1 Part 2


Appendix 2

lobal List Testing Results by ISP

FPT
Appendix 2
VNPT
Appendix 2 B


Appendix 3

Long List Testing Summary

Appendix 3
Figure 25
Summary by Language


Appendix 4

Copy of Block Page from Ho Chi Minh City Cybercafé With VNPT Access

Appendix 4


Appendix 5

Vietnam Background

A. General Description

Vietnam has an estimated population of 84.4 million people (as of July 2006) in Southeast Asia, bordering Cambodia, Laos, and China.172 A French colony from 1884 to 1954, Vietnam achieved unification in 1975 after the withdrawal of U.S. forces from the former South Vietnam.173 It has a Communist political system organized under Leninist principles, and the Communist Party is effectively coextensive with the state.174

Economically, Vietnam has grown rapidly in recent years.175 The state's estimated 2005 Gross Domestic Product (GDP) was U.S. $253.2 billion, with GDP per capita of U.S. $3000 (at purchasing power parity).176 However, Vietnam still ranked 108 out of 173 states in 2005 on the United Nations Development Program's Human Development Index.177 The state has sought growth through limited market-oriented reforms since the adoption of its Doi Moi ("renovation") program in 1986.178 The current leadership will probably maintain pro-reform economic policies.179 The state, though, remains in control of the economy and uses measures such as price controls and centralized planning to achieve objectives.180

B. Political System

Vietnam has an executive branch headed by a chief of state (the president, currently Tran Duc Luong) and a head of state (the prime minister, currently Phan Van Khai).181 The legislative branch is the National Assembly, which appoints the president and the cabinet.182 The president appoints the prime minister.183 Despite formal structures, the Communist Party controls both political and economic policymaking.184 Formal domestic opposition to the Communists is unlikely to emerge in the near future.185 Though nearly 40% of recent candidates for elections to the People's Council were not Communist Party members, candidates must still be reviewed and approved by the state-controlled Vietnam Fatherland Front.186 Political187 and religious188 dissenters risk imprisonment.

Footnotes

1 The OpenNet Initiative is a collaborative partnership of four leading academic institutions: the Citizen Lab at the Munk Centre for International Studies, University of Toronto; the Berkman Center for Internet & Society at Harvard Law School; the Advanced Network Research Group at the Cambridge Security Programme at University of Cambridge; and the Oxford Internet Institute at the University of Oxford. As with all OpenNet Initiative work, these reports represent a large team effort. The work of principal investigators Jonathan L. Zittrain and John G. Palfrey, Jr. on this research report was made possible by a grant from the John D. and Catherine T. MacArthur Foundation's Research and Writing Grants Program of the Program on Global Security and Sustainability. ONI thanks Nam Tran for extensive research assistance on this study.
2 Economist Intelligence Unit, Country Report: Vietnam September 2004 Updater (Sept. 5, 2004) (noting that the government will likely continue to punish cyberdissidents harshly).
3 Article 3, Government Decree No. 55/2001/ND-CP (Aug. 23, 2001), English translation available at http://www.vnnic.net.vn/english/documents/Decree55_2001CP.htm.
4 Southeast Asian Press Alliance, Vietnam readies stricter press laws to rein back aggressive journalists, SEAPA, June 16, 2006, at http://www.seapabkk.org/newdesign/newsdetail.php?No=485.
5 Economist Intelligence Unit, Country Report: Vietnam July 2004 Main Report (July 5, 2004).
6 Economist Intelligence Unit, Country Report: Vietnam June 2004 Updater (June 1, 2004).
7 Southeast Asian Press Alliance, Vietnam readies stricter press laws to rein back aggressive journalists.
8 See Amy Kazmin, Hanoi Steps Up Drive To Weed Out Corruption as "Greed Runs Amok", Financial Times, Dec. 2, 2004, at 11.
9 Human Rights Watch, Vietnam: Torture, Arrests of Montagnard Christians, January 2005, at http://hrw.org/backgrounder/asia/vietnam0105/4.htm#_Toc92870686; see Margie Mason, U.S. requests access to troubled Central Highlands following mass protests, Associated Press, Apr. 13, 2004.
10 Vietnam Closes Website for Publishing BBC Central Highlands Story, Deutsche Presse-Agentur, Apr. 26, 2004.
11 VIR, MPT to tighten the net cyber rules, VietNamNet, June 18, 2005, at http://english.vietnamnet.vn/tech/2005/06/455883/.
12 Dinh Hang, No agency can manage all online news: Deputy Minister, VietNamNet Bridge, Mar. 1, 2005, at http://english.vietnamnet.vn/news/2005/03/384104/ (quoting Deputy Minister of Culture and Information Do Quy Doan).
13 Dinh Hang, No agency can manage all online news: Deputy Minister.
14 See, e.g., GIPI, Promoting Internet Policy and Regulatory Reform in Vietnam Mar. 14, 2004, at http://www.internetpolicy.net/about/20040300vietnam.pdf (stating that "Sites that are considered offensive or contrary to the government's perspective are blocked"); see generally Human Rights Watch, Vietnam: Fledgling Democracy Movement Under Threat, May 11, 2006, at http://hrw.org/english/docs/2006/05/11/vietna13346.htm.
15 Committee to Protect Journalists, Vietnam, Attacks 2005: Asia, at http://www.cpj.org/attacks05/asia05/vietnam_05.html.
16 See Amnesty International, Viet Nam: Rights Denied in Cyberspace, Nov. 25, 2003, at http://www.amnestyusa.org/countries/vietnam/document.do?id=80256DD400782B8480256DE8004DCA93.
17 Committee to Protect Journalists, Vietnam.
18 See, e.g., Internet Use Increases in Vietnam But Remains in Straitjacket, Channel NewsAsia, Dec. 12, 2003; Le Chi Quang, at http://www.hdvnbtdt.org/article.php3?id_article=88.
19 See Censorship, at http://www.flickr.com/photos/noodlepie/63510858/ (Nov. 14, 2005).
20 Vietnam Internet Network Information Center, Vietnam Internet Statistics, at http://www.vnnic.net/english/statistics/index.html.
21 More People Use Internet in Vietnam, Saigon Times Magazine, Dec. 3, 2005.
22 See 65 Cybercafé Owners Reprimanded in Southern Vietnam, Deutsche Presse-Agentur, Sept. 17, 2004. This also makes estimating the number of Internet users in Vietnam difficult.
23 An ONI researcher was able to purchase pre-paid Internet access cards for several providers from multiple locations with ease while in Vietnam.
24 Article 24, Decree No. 55/2001/ND-CP.
25 Dang Thieu, New Vietnamese Search Engine in the Offing, Saigon Times Daily, July 1, 2004 (noting that the most popular Vietnamese search engines are Netnam and Vinaseek, which are having difficulty competing with Google).
26 See Article 7, ? 2, Decision No. 27/2005/QD-BBCVT; Vietnamese Unicode FAQs, at http://vietunicode.sourceforge.net/main.html; see generally United Nations Development Programme, Unicode Implementation in Vietnam, at http://www.undp.org.vn/unicode/tech.htm. TCVN 6909 appears to be a subset of Unicode.
27 See Unicode & Vietnamese Legacy Character Encodings, at http://vietunicode.sourceforge.net/charset/.
28 See Article 3, Decision No. 27/2005/QD-BBCVT; IANA, .vn ? Vietnam, at http://www.iana.org/root-whois/vn.htm.
29 Hoang Hung, VNNIC Open Domain Name Service Office in Da Nang, VietnamNet, July 23, 2005, at http://english.vietnamnet.vn/tech/2005/07/471553/; see generally Articles 6-13, Decision No. 27/2005/QD-BBCVT (describing allocation of .VN subdomains and related procedures).
30 See Article 7, Decision No. 27/2005/QD-BBCVT.
31 See Sumner Lemon, Inaccurate Report Sparks Fears China May Split ?Net, IDG News Serv., Mar. 2, 2006 (describing the second-level Chinese-language domain operating under China's .CN top-level domain).
32 Article 27, Decree No. 55/2001/ND-CP.
33 Article 27, Decree No. 55/2001/ND-CP.
34 Article 27, Decree No. 55/2001/ND-CP.
35 Article 27, Decree No. 55/2001/ND-CP.
36 ONI confirmed this with a knowledgeable source in Vietnam, who remains anonymous as a security precaution.
37 See http://www.vdc.com.vn/index_ns.shtml. VDC is creating an end-to-end IP network using Cisco Systems equipment. See Cisco Systems, VDC Builds Next-Generation Internet Network with Cisco Systems Equipment, Sept. 16, 2004, at http://newsroom.cisco.com/dlls/global/asiapac/news/2004/pr_09-16.html.
38 Decision No. 06/2006/QD-TTg of Jan. 9, 2006, On Forming the Parent Company ? Vietnam Post and Telecommunications Group; see Article I(2)(d), Decision No. 58/2005/OD-TTg of March 23, 2005, Approving the Scheme On Experimentally Forming Vietnam Post and Telecommunications Conglomerate; English translations available at http://www.mpt.gov.vn/english/legal_doc/?thucdon=ld.
39 Decision No. 58/2005/OD-TTg of March 23, 2005.
40 Article 1(2)(c), Decision No. 58/2005/OD-TTg.
41 Article 1(2), Decision No. 06/2006/QD-TTg.
42 Private communication with the authors.
43 Private communication with the authors.
44 Ministry of Posts and Telematics, Operators, at http://www.mpt.gov.vn/english/statistics/operators.asp?thudon=st; see generally VNNIC, Vietnam Internet Connectivity Map, at http://www.vnnic.net.vn/english/statistics/internetmap/Internetmap.htm.
45 VNPT is the largest IXP in Vietnam.
46 See http://www.etc.com.vn/.
47 See Vietnam Internet Network Information Center, Statistics on Internet Subscriber of ISP: July ? 2004, at http://www.vnnic.net/english/statistics/subs/current.htm; see also Toan Viet Ltd. Co., Vietnam Has 1.44Mln Internet Subscribers, Vietnam News Briefs, Aug. 25, 2004 (noting that VNPT has 57.3% of the ISP market share and FPT Telecommunications, at 20.1%, is the only other ISP with more than 8% market share). ONI research inside Vietnam revealed a different number of ISPs depending on the source consulted; it is clear that while this sector is fluid, only a handful of providers have significant market share in reality.
48 More People Use Internet in Vietnam, Saigon Times Magazine.
49 Dinh Hang, No agency can manage all online news: Deputy Minister (quoting Deputy Minister of Culture and Information Do Quy Doan).
50 See, e.g., Reporters Sans Frontières, Internet Under Surveillance 2004: Vietnam, at http://www.rsf.org/article.php3?id_article=10778.
51 Article 13, Decree No. 55/2001/ND ? CP; Vietnam Keeps Lock on Net Access, Wired News, Aug. 27, 2001, at http://www.wired.com/news/politics/0,1238,46332,00.html.
52 See, e.g., Embassy of the Socialist Republic of Vietnam in the United States of America, 1992 Constitution, at http://www.vietnamembassy-usa.org/learn/gov-constitution5.php3.
53 See generally Committee to Protect Journalists, Vietnam.
54 International Telecommunication Union, Vietnam Internet Case Study 15, March 2002, available at http://www.itu.int/ITU-D/ict/cs/vietnam/material/VNM%20CS.pdf.
55 Committee to Protect Journalists, Vietnam.
56 International Telecommunication Union, Vietnam Internet Case Study at 8.
57 Southeast Asian Press Alliance, Vietnam readies stricter press laws to rein back aggressive journalists (citing reporting by Deutsche Press Agentur).
58 Southeast Asian Press Alliance, Vietnam readies stricter press laws to rein back aggressive journalists (citing penalty for revealing state secrets).
59 Southeast Asian Press Alliance, Vietnam readies stricter press laws to rein back aggressive journalists.
60 Southeast Asian Press Alliance, Vietnam readies stricter press laws to rein back aggressive journalists.
61 See, e.g., Article VII, Inter-Ministerial Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT of July 14, 2005, on Management of Internet Agents, English translation available at http://www.mpt.gov.vn/english/legal_doc/?thucdon=ld (specifying roles and responsibilities for various governmental entities in regulating Internet agents).
62 See, e.g., Decision No. 158/2001/QD-TTg of Oct. 18, 2001, ratifying Vietnam Posts and Telecommunications Development Strategy till 2010 and Orientations till 2020, available at http://db.vnpt.com.vn/legals.asp?dataID=401&ID=63.
63 See, e.g., Decision No. 84/2001/QD-BTC of Sept. 5, 2001 of the Minister of Finance governing the collection of charges and fees levied on the allocation and management of Internet addresses and domain names in Vietnam, available at http://db.vnpt.com.vn/legals.asp?dataID=363&ID=64.
64 See, e.g., Decision No. 180/2003/QD-BBCVT defining the functions, tasks, powers, and organizational structure of the Technical and Informatics Center under the Radio Frequencies Department, available athttp://db.vnpt.com.vn/legals.asp?dataID=559&ID=65. The Ministry publishes many of its regulatory documents in English on its Web site at http://www.mpt.gov.vn/english/legal_doc/?thucdon=ld, though it notes the translations are for reference only.
65 See State Eyes Set to Police the Web, The Vietnam Investment Review, Aug. 16, 2004 (describing a "legally binding instruction" from the ministry requiring cybercafé owners to "provide a list of all the websites visited by their customers in the previous 30 days," along with their credit card details or ID card numbers).
66 VIR, MPT to tighten the net cyber rules; Committee to Protect Journalists, Vietnam.
67 VIR, MPT to tighten the net cyber rules.
68 See, e.g., Decision 3586/QD-GCTT dated July 1, 2003 of the President and CEO of the Vietnam Posts and Telecommunications Corporation (VNPT) promulgating the temporary charges of the pre-paid inter-provincial and international IP telephone service, available at http://db.vnpt.com.vn/legals.asp?dataID=547&ID=66.
69 Article 29, Decree No. 55/2001/ND ? CP. The MPT was originally the Department General of Posts and Telecommunications. See generally Ministry of Posts and Telematics at http://www.mpt.gov.vn/english/introduction/?thucdon=in.
70 Article 30, Decree No. 55/2001/ND ? CP. The Encryption Department is charged with managing the coding and encoding of Internet information, including setting national standards. Id. at Article 35.
71 Article 31, Decree No. 55/2001/ND ? CP.
72 Article 32, Decree No. 55/2001/ND ? CP; see also Dinh Hang, No agency can manage all online news: Deputy Minister (quoting Deputy Minister of Culture and Information Do Quy Doan that MCI controls 70 on-line newspapers of Information and Communication Partners (ICPs), along with roughly 2500 other Web sites).
73 Article 33, Decree No. 55/2001/ND ? CP.
74 Article 34, Decree No. 55/2001/ND ? CP (referring to entities designated for preferential rate treatment in Article 4).
75 Article VII, ? 4, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
76 Article 36, Decree No. 55/2001/ND ? CP.
77 Article 37, Decree No. 55/2001/ND ? CP.
78 Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
79 Article I, ? 1, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
80 Article I, ? 3, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
81 Article I, ? 3(c), Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
82 Article II, ? 1, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
83 Article II, ? 5, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
84 Article II, ? 8, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT. This Article references prohibitions set forth in Article 11 of Decree No. 55/2001/ND-CP, which is discussed in Section 2(C)(3) of this study.
85 Article III, ? 2, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
86 Article III, ? 3, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT. Note that legal requirements and implementation may vary ? an ONI researcher did not observe cybercafés requesting or recording this information during a research trip in March 2006 in Ho Chi Minh City.
87 Article III, ? 3, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
88 Article III, ? 4, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
89 Article III, ? 5, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
90 Article III, ? 5, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
91 Article V, ? 6, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
92 Article VI, ?? 1, 4, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
93 Article 2, ? 2, Decision No. 27/2005/QD-BBCVT of August 11, 2005, Promulgating the Regulation on Management and Use of Internet Resources, English translation available at http://www.mpt.gov.vn/english/legal_doc/?thucdon=ld.
94 Article 2, ? 3, Decision No. 27/2005/QD-BBCVT.
95 Dinh Hang, No agency can manage all online news: Deputy Minister (quoting Deputy Minister Do Quy Doan).
96 See generally Article 8, ?1(e), Decision No. 27/2005/QD-BBCVT (requiring electronic newspapers and Web sites registered in the .VN top-level domain to comply with laws regulating on-line publishing and information).
97 Ministry of Culture and Information, Decision No. 27/2002/QD-BVHTT, Issuing the Regulation on management of, and permit granting for the provision of information and establishment of websites on Internet, art. 4 (Oct. 10, 2002), English translation available at http://laws.dongnai.gov.vn/2001_to_2010/2002/200210/200210100001_en/lawdocument_view.
98 Dinh Hang, No agency can manage all online news: Deputy Minister.
99 Decision No. 27/2002/QD-BVHTT.
100 Government Decree No. 55/2001/ND ? CP.
101 Decree No. 55/2001/ND ? CP, art. 1.
102 Decree No. 55/2001/ND ? CP, arts. 3 & 4.
103 Decree No. 55/2001/ND ? CP, art. 6.
104 Id.
105 Decree No. 55/2001/ND ? CP, art. 10.
106 Decree No. 55/2001/ND ? CP, art. 11.
107 Decree No. 55/2001/ND ? CP, art. 22.
108 See, e.g., Decree No. 55/2001/ND ? CP, arts. 41(2) & 41(5)(h).
109 Decree No. 55/2001/ND ? CP, art. 45.
110 Decree No. 55/2001/ND ? CP, art. 14 (defining Internet content providers as "Units providing information service on the Internet"); see Article 12, Decree No. 55/2001/ND (stating that "Internet information service is one kind of Internet application services including newspaper distribution service (broadcasting, television and e- newspapers), distribution of publications on the Internet, and provision of other of e-information news on the Internet").
111 Article 17, Decree No. 55/2001/ND ? CP.
112 Department General of Posts and Telecommunications, Regulation 1.1.3, Circular 04/2001/TT-TBCD (Nov. 20, 2001). Note that the DGPT is now the MPT.
113 Vietnamese Security Ministry Establishes Special Unit to Tackle Internet Crime, BBC Monitoring International Reports, Aug. 4, 2004.
114 See Decision No. 27/2002/QD-BVHTT. See also Vietnam May Set Up Second Internet Police Force, Deutsche Presse-Agentur, Oct. 28, 2003.
115 Vietnamese Security Ministry Establishes Special Unit to Tackle Internet Crime, BBC Monitoring International Reports, Aug. 4, 2004.
116 Vietnamese Security Ministry Establishes Special Unit to Tackle Internet Crime, BBC Monitoring International Reports, Aug. 4, 2004.
117 Ministry of Posts and Telematics, Regulation on Management and Use of Internet Resources, Decision No. 27/2005/QD-BBCVT (Aug. 11, 2005).
118 Chapter 1, Article 2, Regulations on Internet Resources Management and Use, Decision 92/2003/QD-BBCVT (May 26, 2003). English translation available at http://www.mpt.gov.vn/english/legal_doc/?op=3&thucdon=vb&id=VB3110552965. "Internet resources" are defined to include domain names, Internet Protocol (IP) addresses, Autonomous System Numbers, and other Internet names and numbers. Id. at Chapter 1, Article 5. This decision replaces Decision 92/2003/QD-BBCVT of the Ministry of Posts and Telematics promulgating Regulation on Internet Resources Management and Use (May 26, 2003), which prohibited "use of Internet resources for purposes against the Socialist Republic of Vietnam, disturbing security, economy, social order and safety and violating the national habits and customs."
119 Vietnam Steps Up Control of Internet, Channel NewsAsia, July 21, 2004.
120 65 Cybercafé Owners Reprimanded in Southern Vietnam, Deutsche Presse-Agentur, Sept. 17, 2004.
121 Ministry of Posts and Telematics, Directive No. 06/2004/CT-BBCVT (May 7, 2004).
122 Directive No. 06/2004/CT-BBCVT (May 7, 2004), Ministry of Posts and Telematics.
123 Directive No. 06/2004/CT-BBCVT (May 7, 2004), Ministry of Posts and Telematics.
124 Directive No. 06/2004/CT-BBCVT (May 7, 2004), Ministry of Posts and Telematics.
125 Directive No. 06/2004/CT-BBCVT (May 7, 2004), Ministry of Posts and Telematics.
126 Directive No. 07/2004/CT-BBCVT (July 19, 2004), Ministry of Posts and Telematics. English translation available at http://www.mpt.gov.vn/english/legal_doc/?op=3&thucdon=vb&id=VB2460435021.
127 Decree No. 55/2001/ND ? CP, art. 20.
128 Article 20, Decree No. 55/2001/ND ? CP.
129 Article 20, Decree No. 55/2001/ND ? CP.
130 Article 20, Decree No. 55/2001/ND ? CP. See also Vietnam Tightens Control Over Cybercafés, South China Morning Post, July 21, 2004 (reporting that the directive permits ISPs to "suspend contracts with cybercafés that allow clients to access pornographic sites or those deemed to threaten national security").
131 Article 20, Decree No. 55/2001/ND ? CP.
132 Article 20, Decree No. 55/2001/ND ? CP.
133 Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
134 Article I, ? 3, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
135 Article V, ? 3, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
136 Article V, ? 4, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
137 Article V, ? 9, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT.
138 State eyes set to police the Web, The Vietnam Investment Review, Aug. 16, 2004 (noting also that "Internet cafés are not allowed to offer ?unhealthy games and discs'").
139 ONI thanks Thong and Nhung Nguyen for expert translation assistance.
140 These regulations reflect the requirements set forth in Article II, ? 8, Joint Circular No. 02/2005/TTLT-BCVT-VHTT-CA-KHDT, and Article 11, Decree No. 55/2001/ND-CP.
141 See VIR, MPT to tighten the net cyber rules (noting that a circular promulgated in 2005 by several ministries requires café owners to warn users that they must not access pornography and to position computer monitors so that staff can supervise what content customers view).
142 State eyes set to police the Web, The Vietnam Investment Review.
143 State eyes set to police the Web, The Vietnam Investment Review.
144 Reuters, Internet Cafés Shut in Vietnam for Porn, Politics, Express India, Nov. 23, 2004, at http://www.expressindia.com/fullstory.php?newsid=36235.
145 See No Agency Can Manage All Online News: Deputy Minister, VietNamNet, Mar. 1, 2005, at http://english.vietnamnet.vn/news/2005/03/384104/.
146 No Agency Can Manage All Online News: Deputy Minister, VietNamNet, Mar. 1, 2005, at http://english.vietnamnet.vn/news/2005/03/384104/.
147 No Agency Can Manage All Online News: Deputy Minister, VietNamNet, Mar. 1, 2005, at http://english.vietnamnet.vn/news/2005/03/384104/.
148 No Agency Can Manage All Online News: Deputy Minister, VietNamNet, Mar. 1, 2005, at http://english.vietnamnet.vn/news/2005/03/384104/.
149 No Agency Can Manage All Online News: Deputy Minister, VietNamNet, Mar. 1, 2005, at http://english.vietnamnet.vn/news/2005/03/384104/.
150 An ONI research team member performed testing in Ho Chi Minh City during March 2006.
151 See Human Rights Watch, Repression of Montagnards, at http://www.hrw.org/reports/2002/vietnam/; see also Police Disperse Vietnam Protest, BBC News, Apr. 11, 2004, at http://news.bbc.co.uk/2/hi/asia-pacific/3617305.stm.
152 See, e.g., Didier Lauras, Controversial Border Agreement at Heart of Cyber Dissident Trial, Agence France-Presse, Dec. 30, 2003.
153 See United States Department of State, International Religious Freedom Report 2004: Vietnam, at http://www.state.gov/g/drl/rls/irf/2004/35433.htm (Sept. 15, 2004) (noting that respect for religious freedom for independent Buddhist groups in Vietnam has deteriorated).
154 See BBC Monitoring International Reports, Vietnam Asks ISPs to Block Foreign Sites Offering Illegal Phone Services, Nov. 16, 2003 (quoting the director of Vietnam Datacommunication Company as saying that the company was "increasing the use of firewalls to block the penetration of illegal Internet web sites").
155 See Congressional Budget Office, H.R. 1587: Vietnam Human Rights Act of 2004, at http://www.cbo.gov/showdoc.cfm?index=5639&sequence=0.
156 See United States Department of State, International Religious Freedom Report 2004: Vietnam (noting that Cao Dai adherents resisted the Communist movement until 1975, contributing to their subsequent persecution).
157 Committee to Protect Journalists, Vietnam.
158 See Austin Ramzy, Milestones ? Events of the Week, Time Asia Magazine, Jan. 12, 2004, at http://www.time.com/time/asia/magazine/article/0,13673,501040112-570315,00.html.
159 See generally Human Rights Watch, Vietnam: Fledgling Democracy Movement Under Threat (listing dissidents); Committee to Protect Journalists, Vietnam.
160 See Appendix 2 for categorized results of the global list testing for both ISPs. To provide comparable results across multiple country studies, the majority of the sites in our global list have content only in English.
161 The remaining sites on the global list for each ISP were not available for reasons unrelated to filtering.
162 The number of sites tested for each ISP varies slightly as some sites were not accessible during testing or returned other errors unrelated to filtering, such as transient network errors.
163 The Open Directory Project is available at http://www.dmoz.org/. Dmoz classified over 4 million sites into 590,000 categories as of July 12, 2006. The complete taxonomy can be downloaded at http://rdf.dmoz.org/.
164 As above, the difference in the number of URLs tested results from Web sites that were inaccessible for reasons unrelated to filtering, such as network errors.
165 See Appendix 3 for a summary of our analysis.
166 See About Us, International Vietnamese Youth Network, at http://www.lenduong.net/article.php3?id_article=7746 (last updated Apr. 24, 2004).
167 See Vietnamese Professionals Society at http://www.vps.org/article.php3?id_article=191.
168 See generally Religion, Montagnards ? Their History and Culture, at http://www.culturalorientation.net/montagnards/vrelig.html (last updated Feb. 18, 2004).
169 The Vietnamese-language search term used was "D? Lu?t Nh?n Quy?n Cho Vi?t Nam".
170 See Appendix 4 for an image of the VNPT block page collected at a cybercafé in Ho Chi Minh City, Vietnam.
171 The A records, as listed in the DNS, map a URL (such as www.ykien.net) to the IP address of the server that hosts the Web site for that URL. See P. Mockapetris, Domain Names ? Implementation and Specification ? 3.4.1 (Nov. 1987), available at http://www.ietf.org/rfc/rfc1035.txt. The user's browser then attempts to connect to the Web site at that IP address and to retrieve the requested material.
172 CIA, The World Fact Book -- Vietnam, at http://www.cia.gov/cia/publications/factbook/geos/vm.html.
173 CIA, The World Fact Book -- Vietnam.
174 Economist.com, Country Briefings ? Vietnam, at http://www.economist.com/countries/Vietnam/profile.cfm?folder=Profile-FactSheet.
175 International Telecommunication Union, Vietnam Internet Case Study at 2.
176 CIA, The World Fact Book -- Vietnam.
177 United Nations Development Program, Human Development Report 2005, at http://hdr.undp.org/reports/global/2005/pdf/HDR05_complete.pdf.
178 International Telecommunication Union, Vietnam Internet Case Study at 1.
179 Economist Intelligence Unit, Country Report: Vietnam November 2004 Updater (Nov. 9, 2004).
180 International Telecommunication Union, Vietnam Internet Case Study at 1.
181 CIA, The World Fact Book -- Vietnam.
182 International Telecommunication Union, Vietnam Internet Case Study at 3.
183 CIA, The World Fact Book -- Vietnam.
184 Economist Intelligence Unit, Country Report: Vietnam September 2004 Updater at 1-3.
185 Economist Intelligence Unit, Country Report: Vietnam September 2004 Updater at 1-3.
186 Economist Intelligence Unit, Country Report: Vietnam April 2004 Main Report (Apr. 1, 2004).
187 See, e.g., Amy Kazmin, Outcry at jailing of Vietnamese dissident, Financial Times, Aug. 2, 2004, at 2 (describing Vietnam's decision to sentence ailing dissident Nguyen Dan Que, who "wrote an article about Vietnam's strict control over information and the media, and posted the article on the internet," to 2.5 years in prison).
188 See, e.g., Protestant dissident sentenced to three years in Vietnam, Deutsche Press-Agentur, Nov. 12, 2004 (describing the sentencing of Reverend Nguyen Hong Quang to three years in prison for "opposing officials who are carrying out their duty").