On September 17, 2012, the government of Pakistan banned YouTube access throughout the country. An anti-Islam video that had been deemed blasphemous was the cause of immense anger in the country, so the government blocked YouTube. The official reason given is that it was conducting its moral duty of protecting its citizens from offensive content. The government approached Google and demanded the removal of the video. Google did not comply and, consequently, YouTube has been blocked in Pakistan ever since (except for 3 minutes).
This issue raises bigger questions about the purpose served by Internet censorship, particularly in a networked world:
- Does the Pakistani government really believe requesting Google to remove the video will help prevent the content from being viewed in Pakistan? Copies of the video and similar offensive content will always spring up. Nearly 8 years of content are uploaded to YouTube every day!
- Does the Pakistani government really believe blocking YouTube will help prevent the content from being viewed in Pakistan? There are dozens of alternatives to YouTube. The government has not blocked them though it does not have the means of verifying whether the offending video or similar videos exist there. Moreover, people will find ways to view the content if they want to. It is just an inconvenience to them. The content can still be viewed in Pakistan.
- Does the Pakistani government really feel that it is fulfilling its moral duty to its citizens with this act of censorship? On the surface, their actions reek of hypocrisy. However, when one digs deeper, this is a much more complicated issue. Let me return to this point later.
It appears as though the real reason behind this ban is to counter protests and violence within the country. In 2010, Pakistan blocked YouTube for 6 days over fear of the repercussions from the ”Everyone Draw Mohammad Day” competition posted by a Facebook user. Pakistan is a country stricken with violence, and bomb blasts that are triggered by remote controls to cellphones are becoming increasingly common. To counter them, the government often shuts down mobile service networks on certain occasions when higher numbers of attacks are expected. It is possible that such temporary measures do indeed save lives because there seem to be fewer remote-controlled bomb blasts during the blackout periods. However, other forms of attacks continue, as do remote-controlled bomb blasts when mobile service is restored. Clearly this is, at best, an extremely limited deterrent to the violence. Bans are just band-aids. On the topic of internet censorship, it should be noted that Pakistan pales in comparison to some other countries.
The Silencing Approach – Internet Blackout of Egypt
It doesn’t get any more drastic than this. During the Arab spring, 3500 BGP prefixes were removed when the Egyptian government realized how influential Twitter and Facebook were to the political uprising. Approximately 90% of Egypt was inaccessible in the last week of January 2011. For one week, dial-up access from landlines once again became common.
The Walled Garden Approach – Iran and China
Iran and China prefer a more subtle way – monitor the activities of its citizens. In September 2012, a study funded by the University of Pennsylvania identified many domains with DNS A records of both public and private (RFC 1918) addresses in Iran. This evidence of ‘dual-stack’ networks corroborates claims of an intranet being developed in Iran. In December 2012 Iran announced its own version of YouTube – a video-sharing site called Mehr. China built Baidu as an alternative search engine to Google to complement its Great Firewall. Obviously, these countries claim to be acting in their own national interest for the safety and security of their own citizens. It is widely believed that the USA and Israel were behind the Stuxnet worm that grew out of the Olympic Games cyberattacks on Iran.
The Pakistani government also claims it is doing its duty by protecting its citizens from offensive content. However, they can only protect from the repercussions of the content. Technology is just a tool, and repressing it is a stop-gap measure. Bans are not sustainable solutions. The government has done nothing to address the root of the problem, which is to curb the intentions behind such violence. That is better left for a separate post. So, granted that not all countries provide freedom of speech as a constitutional right as in the United States, let us return to the question of whether the Pakistani government truly believes it is fulfilling its moral duty to conduct internet censorship.
Prohibitionism and Misaligned Incentives
In 1919, the Volstead Act was enacted in the United States to prohibit the sale, manufacture, and transportation of alcohol. That it completely backfired, and was actually a boon to organized crime, is attributed to the nature of human beings: we are attracted to that which is taboo. If the Pakistani government really wants to protect its citizens from offensive content, it should remove its filters, because they only serve as challenges for people to come up with creative ways to circumvent them. The YouTube ban hasn’t dented the resolve of resourceful Pakistanis who have become experts in leveraging IP anonymizer services. Indeed, a search for how to access youtube in pakistan yields 41 million Google results!
A Truly Networked World and the Identifiable Victim Effect
The Pakistani government has frequently sought to ban or boycott products or websites that are considered blasphemous or offensive. So whereas one website may be considered perfectly friendly (perhaps a religious site) and promoted by the government, another site that is considered offensive (perhaps a pornographic site) could be hosted by the same public cloud provider in the same data center. How does the government react to such information? Would they approach the cloud provider and make similar ludicrous requests to move the webserver(s) that hosts the friendly site to a different network? Would they choose a different provider? Now let’s complicate things even further by bringing virtualization into the picture. It is very possible that the desirable content promoted by the government and the content that is considered offensive by the government are hosted on the same physical server. How would the government feel if they were to find out that the two VMs share the same CPUs, same physical NICs, and could be accessing storage on the same array? Religious content and porn could be sharing the same wire! How does that taint things? Is sharing bandwidth with an offensive VM considered blasphemous? Where do you draw the line?
Obviously, these are all just hypothetical questions (one of the goals of cloud security is for tenants to be completely oblivious of one another). Nevertheless, they are questions that the Pakistani government should be asking itself if it truly believes it is acting as the moral savior. Internet censorship is a slippery slope.
It is more than likely that the Pakistani government probably doesn’t think about this too much, let alone care about it. The YouTube ban is a prime example of the Identifiable Victim Effect whereby action is only taken what is observed and identified. In other words, ignorance is bliss!