Google entered China in 2006 with the belief that overall it could increase access to information for Chinese users, even if it meant censoring some results. But in the face of increasingly strident government demands to limit free speech online, Google faces the difficult task of deciding whether the balance between human right risk and opportunity has tipped to the point where they can no longer operate ethically in China. And further to its credit, Google is asking the government to respect the human rights of Chinese users. It is precisely this kind of engagement with human rights issues that all companies should incorporate into their business operations.

Putting aside Google’s extraordinary announcement for a moment, this is second round of embarrassment the Chinese government has faced in recent weeks. Despite its efforts to quietly shutter the beleaguered Green Dam initiative, last week, CYBERsitter, a US-based web filtering software company, filed a lawsuit against the People’s Republic of China, two Chinese software makers involved in the development of Green Dam, and seven major computer manufacturers that had agreed to install the software, seeking US $2.2 bn in damages. The suit alleges copyright infringement, it claims the software makers copied code from CYBERsitter’s software and conspired with the Chinese government and computer manufacturers to distribute more than 56m copies of Green Dam. CYBERsitter also alleges that it has been hit by thousands of cyber attacks since it first raised copyright concerns in the media last year, including some that may have come from inside China. These claims take on a new light following Google’s revelation that the company has been the target of sophisticated cyberattacks.

This latest Green Dam episode also raises interesting questions about how filtering software developed in the US — with its strong tradition of protecting free expression online — can also be used by governments to actually restrict speech abroad. Researchers over at the OpenNet Initiative have documented how filters developed by US companies are now being used by authoritarian regimes to restrict expression and access to information online, often to stifle dissent.

The Chinese government ostensibly promoted Green Dam as an effort to protect children from sexually explicit material online, a goal that many governments share. In the US, free expression advocates argue that the best way to protect children online while minimising the impact on protected speech is to empower parents to educate their kids and, if they choose, use tools like filters to control what content their children can access online. The US has a robust and competitive market for filtering software (including CYBERsitter) aimed at helping parents protect their children from a variety of harms online but the notion of user control is key.

In the case of Green Dam, however, the Chinese government mandated the use of one particular filter on all computers, and users quickly discovered that the software went much further than merely filtering pornography –– it also blocked content on “politically sensitive” topics like democracy and Tibet, monitored individual computer behaviour, and created serious security issues. Green Dam was revealed to be the next step in the government’s effort to decentralise internet repression. In the face of widespread user discontent the Chinese government was forced to scrap the project.

But what would have happened if the Chinese government had approached CYBERsitter (or any other US company) for a license to adapt their software for Green Dam or simply for use by state-controlled ISPs? Recent controversies over human rights in the ICT sector have shown it is not enough for a filtering software company to say that they have no responsibility for foreseeable use of their products in violations of human rights. Customers, users, human rights groups, shareholders, and their home governments all have a stake in ensuring ICT companies promote free expression and avoid complicity in human rights violations. Meeting this responsibility involves examining the risk that the company’s software could be used to stifle expression if sold to governments (or governmental proxies like state-controlled ISPs) with a history of doing so. Companies must then craft strategies to minimise the foreseeable negative impact on human rights — including choosing not to sell certain software solutions to authoritarian states. Software companies must develop proactive and systematic ways of addressing these risks.

Governments also have a role in setting the standard for corporate responsibility for software companies. Western governments could subject filtering software to export control where there is a risk that the software could abused by repressive regimes. In a perfect world, this might mean that the US government could forbid the sale of network-based filtering software to governments with a history of repressive internet censorship, while still allowing the sale of client-side software that is fully controlled by the user.

However this solution would not change the reality that in many closed societies users may never have access to filtering software that hasn’t been altered or configured by the government for political purposes. Updating export control policies to prevent filtering software being exploited as a censor is just one piece of the larger strategy for advancing internet freedom. The belief that freedom is best advanced when tools remain in the hands of users can go both ways: As we saw in 2009 in Iran and elsewhere, expression on web 2.0 platforms is vital to many democratic and human rights movements all around the world. Governments should be using all the diplomatic and economic tools at their disposal to promote such tools – including actually promoting the export of software and services that enable free speech and facilitate the flow of information in the era of YouTube and Twitter.

Cynthia Wong is Staff Attorney and Plesser Fellow at the
Center for Democracy & Technology