Last month the OpenNet Initiative published a report that shines light on one of the more sensitive business practices of Western Internet security and filtering companies. These companies – including McAfee (an Intel subsidiary), Websense, and Netsweeper – promote their filtering technologies in the West as tools for parents and schools trying to shield children from online pornography and employers looking to maintain a professional work environment. But they also appear to make their software and URL categorization services available to state-run ISPs and telecoms in Middle Eastern and North African countries, such as Bahrain, UAE, Qatar, Oman, Saudi Arabia, Kuwait, Yemen, Sudan, and Tunisia. These ISPs and telecoms, and the governments behind them, use the software to filter out Internet content that they don’t want their citizens to see.
What content? Well, depending on the Western software company, any of the millions to billions of websites that the company has categorized. And the categories, of which multiple companies boast that they have more than 90, range from porn and violence, to dating and filesharing, to politics, religion, and even anonymizers. All the repressive regime has to do is to buy the software, pay the Western company to maintain the database of categorized websites, click the check boxes next to the categories of sites that it doesn’t want its people to access, and viola, the Western company has commercialized censorship. As the report puts it, “This is not simply a case of a general purpose, neutral tool being used for an end not contemplated by its maker. The filtering products of today engage in regular communications with their makers, updating lists of millions of websites to block across dozens of content categories, including political opposition and human rights.”
The report illustrates how the categorized lists these companies maintain tend to be overinclusive – after all, a governmental customer is unlikely to care if more speech is censored than necessary as long as nothing that it doesn’t want its citizens to see gets through. Furthermore, to give a repressive state the flexibility it needs to oppress effectively, most Western companies also allow their governmental customers to create user-defined lists of sites to filter, in case there is additional content that the government wants to block. Finally, some combination of the Western companies and the governments who use their products has recently moved to obscure attribution of filtering to these products, so citizens – and groups like the OpenNet Initiative – have a hard time determining who is allowing their government to censor the Internet.
It doesn’t have to work this way. Western companies don’t have to sell their filtering tools to repressive regimes – or any government or state-run ISP. They could limit customers to individuals and private employers. Moreover, they don’t need to maintain lists of categorized sites at all. And even if they want to keep lists of violent or pornographic sites for legitimate users, classifications such as “politics,” “religion,” and “privacy” are inexplicable unless the Western company is actively trying to help its governmental customers muzzle speech, and inexcusable then. Therefore, at a minimum, the Western companies could get rid of many of their categories.
Risks would still exist. Governments could steal the technology, as Iran may have done with McAfee’s SmartFilter. And in certain cases, repressive regimes could adapt free software developed for innocuous purposes to filter their citizens’ Internet. These risks – and others – may be sufficient to counsel against supplying anyone with any tool that can be repurposed for state-level censorship. But at the very least, Western companies shouldn’t be continuously complicit in government Internet censorship by selling repressive regimes the software and regularly providing them with updated lists of sites to filter.
It’s remarkable how brazen these Western filtering companies are. For example, one American company, Websense, has an explicit policy not to facilitate government censorship, except to restrict pornography. But among its nearly one hundred classifications listed in the report are such categories as “Advocacy Groups,” “Traditional Religions,” “Political Organizations,” and “Educational Institutions.” Perhaps Websense can articulate a legitimate reason for these categories, but it seems a stretch to relate them to “Adult Content,” which is a separate category in any case.
Another company, Netsweeper, is apparently perfectly willing exploit the freedom of foreign peoples by selling its software to government-backed ISPs looking to “block inappropriate content to meet government rules and regulations ‘based on social, religious or political ideals.’” Meanwhile, McAfee remains mum on how its relationship with repressive governments plays into its business conduct and ethics policy.
In an online world where we condemn oppression of a single netizen as cyberbullying, what do we call the conduct of Western companies that collude with governments to oppress an entire citizenry? Cyberrepression? And should companies that ostensibly exist largely to give parents the control needed to shield children from harmful Internet content be surprised if the government that created them exerts another form of parental control – the kind that parents use on poorly-behaving children with no self-control – by regulating the companies’ own asocial behavior? After all, if corporations have rights and obligations based on the legal fiction of corporate personhood, then these companies are the all-too-real sociopaths of the corporate world.
Even better, customers in Western countries can send a free-market message to these companies without having to resort to a regulatory intermediary: such duplicitous behavior – marketing software in the West as a tool to empower parents and businesses but in the Middle East and Africa as a tool to enervate a state’s citizenry – isn’t acceptable. We shouldn’t buy software that’s supposed to protect if its maker also sells it as a means to abuse.