RConversation

ICANN, civil society, and free speech

Gordon Crovitz's Op-ed in the Wall Street Journal about the relationship between ICANN and the future of free speech quotes me a couple of times. An extended excerpt:

Rebecca MacKinnon, a Web researcher writing a book about lessons from China on Internet freedom, praises Icann for being influenced by nongovernmental groups, not just governments. “The U.N. model of Internet governance is highly unsatisfactory from a human-rights and free-expression point of view for obvious reasons,” she told me. “The Chinese and the Iranians and various other authoritarian countries will insist on standards and rules that make dissent more difficult, destroy the possibility of anonymity, and facilitate surveillance.”

Up to now, governments have been largely hands-off. An amusing example is the dispute over the domain www.newzealand.com. The queen of England, “in right of her Government in New Zealand, as Trustee for the Citizens, Organizations and State of New Zealand,” brought an action in 2002 against a Seattle-based company called Virtual Countries Inc. that had registered the Web address. The queen argued that her antipodean country should have control over its own .com name. This may sound reasonable, but she lost. New Zealand had to buy the .com address for $500,000.

Will governments like China’s be as philosophical about Internet domain decisions they don’t like?

Countries such as China, Russia and Iran have long argued that it’s wrong for Icann to report to the U.S. government. Any alternative to the light control exerted by the U.S. government could put the Web on a slippery course toward more control. This is one reason efforts by these countries to politicize Icann have failed in the past.

“I think the question here is not about which governments have the moral right to lead Internet governance over others,” Ms. MacKinnon argues, “but about whether it’s appropriate that Internet governance should be the sole province of governments, many of which do not arguably represent the interests of Internet users in their countries because they were not democratically elected.”

Crovitz emailed me last week when he was researching his column. I was somewhat more critical of ICANN's status quo in our exchange than in the quotes he ended up using. Below are my full answers, emailed to him on Thursday (hyperlinks added for people not familiar with some of the names, events, and terms to which I refer without explanation). If you are not familiar with ICANN acronyms and developments you may want to read this post and this post first.

Q: Why has China re-engaged with ICANN and what are Beijing's main concerns?

A: First of all, the issue of how Taiwan is to be named and treated is now resolved, and given the improvement in cross straits relations, sitting around the same table with the Taiwanese delegation has ceased to be an insurmountable issue. Secondly, ICANN is at a major juncture as it prepares to roll out internationalized TLD's, which means that URLs, completely in Chinese, before and after the dot, will become internationally possible for the first time ever in 2010. There are a bunch of technical issues with the way IDN TLD's get implemented, including whether it will be possible to have less than three characters in IDN gTLDs, etc., etc. (And given that most words in Chinese are 2 characters, that's a big deal which the mainly Western staff has a hard time getting their heads round.) There is also a long list of outstanding, unresolved details regarding the "new applicant guidebook" which will be finalized later this year: requirements for who can apply to run a new TLD, how much they'll be charged, who has the right to object to the use of certain brand names or place names or other names that certain cultures feel they have claim to, etc. The Chinese government and Chinese companies clearly have an interest in the details of how these things get implemented because it will impact how the Chinese language Internet can evolve internationally in the years to come. So Beijing really can't afford to be disengaged from ICANN if it wants to have a role in shaping these things.

Another reason China disengaged from ICANN was that during the WSIS process it and a number of other countries (Iran, Brazil, the EU) were pushing to have internet governance moved out of ICANN to the ITU or some other UN-like or UN body. However at WSIS in 2005 it became clear that there wasn't sufficient international consensus for that to happen, human rights groups (including HRW) were vocal about the free speech implications, and the U.S. insisted on maintaining the status quo. A compromise was the formation of the IGF- a yearly multistakeholder forum where people from all around the world - governments, corporations, and civil society, get together and try to work out what the main issues are on Internet governance and where things shoudl go from here. China has declared the IGF to be a waste of time that should be discontinued. It argues that that Internet governance should be left to governments, who adequately represent the interests of internet users, and that the multistakeholder governance model (which ICANN also uses, but as part of actual decision-making structures unlike the IGF which is only a talk shop) is ineffective and unnecessary.

Q: What concerns should others have about China's re-engagement?

A: China's re-engagement with ICANN is actually a good thing, as long as ICANN maintains its multistakeholder, consensus-based decisionmaking structure, and as long as the Government Advisory Council (which China rejoined) is not given excessive powers disproportionate to other stakeholder groups within ICANN. Chinese registrars and registries have remained engaged throughout ICANN's life, actually. It's just that the government wasn't engaged. The issue is whether China will feel that its interests (many of which are reasonable, like properly serving the linguistic and technical needs of Chinese Internet users and making it possible for Chinese companies to operate on an even playing field with Western ones) are adequately served within the ICANN structure, or whether it ultimately concludes that ICANN can't ultimately serve its interests, in which case they will go back to advocating its total dismantlement - or lead the way in splitting the root, taking the non-English speaking developing world with them.

The key in my view is to ensure that Chinese non-governmental civil society interests, and the interests of Chinese speakers not living in the PRC, can be heard and upheld by ICANN as it works out the shape of the global multilingual Internet. Whether or not that happens depends in part on whether non-PRC and Chinese non-governmental groups recognize the importance of engaging with ICANN, and get the financial support they need to engage, attend its meetings, and get involved with GNSO council through which at-large users and non-commercial voices can voice their concerns and seek representation. Right now Chinese civil society voices are absent from ICANN. Nobody's blocking them from being there necessarily - they're just not there. I guess mainly because very few people in the human rights/free expression space are paying any attention to ICANN if they even know it exists..

Q: The planned IDN/gTLD expansions could empower groups that might not be in political favor by giving them their own extensions, etc.. Does the governance structure of ICANN, with ultimate authority resting in the U.S. Dept. of Commerce., create an opportunity for greater rights to free expression? How has ICANN and Commerce been doing with this reponsibility?

A: From everything I have observed, the U.S. government's relationship with ICANN is extremely hands-off. That may be in part because ICANN isn't doing anything that goes against U.S. interests. But ICANN's decisionmaking structure is very complex. It's not like they're taking orders from Gary Locke and his people. They have a board and several layers of councils and constituencies who feed into the board's decision-making process, in addition to a large number of paid staff, and a CEO. The engineers who actually make the DNS work have a lot of power. So do the commercial registrars and registries, and other groups. Then there is an at-large user constituency and a non-commercial user constituency, both of which feed into the GNSO council which then feeds into the board. Etc. I do get the sense that the primary criteria for decision making has to do with what will keep the global internet working smoothly and what will make the domain name system globally inter-operable, rather than what the U.S. government wants.

I can certainly see why everybody except Americans think it's unfair, however, that the U.S. has ultimate power over ICANN. Especially at a time when Americans are no longer the biggest user group on the internet,  we're shrinking fast as a percentage of overall global internet users. Chinese are now the biggest user group and they're only going to get bigger. Whatever I think of the Chinese government, I can understand how the situation looks to non-Americans. Insisting that the U.S. should keep the keys to ICANN is like saying that only white male Protestant land-owners from Virginia deserve the right to have the final say over how the U.S. is run. In the long run insisting on the status quo is a very un-democratic position to hold, if we are actually going to be intellectually honest.

That said, the UN model of internet governance is highly unsatisfactory from a human rights and free expression point of view for obvious reasons: the Chinese and the Iranians and various other authoritarian countries will insist on standards and rules that make dissent more difficult, destroy the possibility of anonymity and facilitate surveillance. No good. So what to do?

Ultimately, I think the question here is not about which governments have the moral right to lead Internet governance over others, but about whether it's appropriate that Internet governance should be the sole province of governments, many of which do not arguably represent the interests of Internet users in their countries because they were not democratically elected. In addition, I don't trust any government or any collection of governments - even if it were restricted to mature democracies - to uphold my freedom of speech and civil liberties. I don't think the U.S. government has been doing such a stellar job lately, and it too is placing increasing priority on security and fighting crime at the expense, I believe, of free speech and privacy. Sure we need security, but I don't think the USG had the balance right under Bush and I think Obama will disappoint the human rights/free speech community as well. Which is one reason why I think that having a multistakeholder model of Internet governance - in which both civil society and a range of private interests are able to have an equal role in decisionmaking alongside government representatives - is extremely important. There are too many constituencies across the global Internet whose interests cannot not be adequately represented, and whose rights cannot be adequately protected, by traditional geopolitical constructs.

So in my ideal scenario, ICANN's agreement with Commerce gets renewed in September but acknowledging the reality that ICANN's governance model has to evolve into something both more international and more multistakeholder-based over the longer term. Because if it doesn't, and if large parts of the world led by China eventually decide they want to create an alternative DNS not governed by ICANN because ICANN's Internet fails to suit their needs, they can and will do that. And that would be much, much worse. ICANN only has any power at all because everybody in the world has agreed to use its DNS. 

Meanwhile efforts need to be made to help non-governmental civil society constituencies from Asia, the Middle East, Latin America, the former Soviet Bloc, and Africa gain a stronger foothold and voice within ICANN's multistakeholder structure. So that when the inevitable happens and the U.S. moves over to share power with other governments, as is only reasonable as Americans become a shrinking minority of Internet users, there will be plenty of other stakeholders involved with the process making sure that the governments who weren't elected and who would prefer to suppress dissent are not able to hijack the Internet's future as they've hijacked certain UN bodies.

***

So ends the e-mail exchange. One thing I did not get into with Crovitz are my concerns about whether ICANN's multi-stakeholder model is truly effective, and whether it is currently doing enough to listen to the needs of Internet users in the developing world and people who stand to benefit most from the forthcoming expansion of IDN gTLD's. Right now, I don't think ICANN is doing an adequate job at incorporating the concerns of non-commercial, civil society interests into its decision-making process.

If ICANN is serious about serving the needs and interests of all of the world's Internet users - not just the rich Western ones - they need to find a way to strengthen, support, and empower ICANN's non-commercial users. As the Internet expands to become a largely non-English place, the concerns of non-English speakers, people from the developing world, and people who are not represented by democratic governments become much more important.  ICANN needs to find a way to make sure these people's concerns are heard and rights are protected. Right now, it's a fact that the ICANN community is not adequately seeking out or incorporating the views of the people who are fast becoming the majority of Internet users on the planet.

What does the proposed New Applicant Guidebook mean for Iranian dissidents? For Tibetans and Uighurs? For native speakers of Urdu, Bangla, and Malagasy? How might the recommendations proposed in the IRT report on trademark protection play out for potential domain registries and registrants in the developing world? I have seen no evidence when I was at Sydney that these questions have been seriously explored. If ICANN is serious about its stated mission, shouldn't they be?

The board hasn't made nearly enough effort to seek answers to these questions as they prepare to expand the gTLD space and roll out IDNs - assuming they really do care about the answers, which frankly I think they have failed to prove. On the other hand, in my humble opinion, the Non-Commercial User Constituency (NCUC) hasn't done nearly enough to play the role it could be playing.

The NCUC is currently struggling with ICANN staff over their governance charter. NCUC members had engaged in a consensus process to create a charter that they felt would best enable them to represent the interests of non-commercial Internet  users around the world. ICANN staff has rejected that charter and is instead asking them to adopt a top-down governance structure, through which they would only be able to directly elect three out of the six NCUC representatives on the GNSO council. I won't get into all the details but you can read the discussion of the issue here and here, among other places.

Based on what various people at ICANN have told me, there appear to be concerns amongst some members of the ICANN board that the NCUC is dominated by a certain group of American lawyers and academics, who for some reason they don't like. There are a lot of personal animosities in both directions that seem to go way back to when ICANN was founded. Who gives a flying fart about those fights beyond the people involved? Everybody involved with ICANN needs to put the past behind them and figure out how to prove to the world that it can indeed be an effective, truly multi-stakeholder organization that represents the interests of the world's Internet users.

Until recently, the NCUC (which in my observation has in fact until now been run mainly by American and European academics and lawyers), has not had the depth of expertise and experience among its membership - and leadership - to be truly helpful on matters of greatest concern to the developing and non-English speaking world. Nor has the At-Large constituency - the other potential vehicle for individual user interests seemed to be very useful in that regard. But the NCUC has in the past few months undergone a big recruitment drive and has brought in a lot of new members from around the world. I'm optimistic that with its new bottom-up structure (if ICANN allows it) the concerns of a new, more diverse membership will be able to drive the constituency's work at ICANN in the future. I'm also optimistic that if members of civil society groups around the world view the NCUC as an effective way to engage with ICANN, they will make greater effort to participate. It's time for ICANN to prove that its multi-stakeholder bottom-up process, which new CEO Rod Beckstrom has been bragging about, is not a sham.

ICANN and Free Speech

Upon being appointed as ICANN's new CEO in Sydney, Rod Beckstrom gave a rousing speech in which he stressed the vital importance of free expression on the Internet:

I believe in a world where every human being has the ability to communicate with other human beings openly and freely.

He then evoked recent events in Iran:

... what more poignant event do we need than what we've seen in the last few weeks, when people have sought to share their sides of a story -- both sides of a story, without taking a position -- in what's going on in Iran around a democratic election, as people have tweeted, Facebooked, YouTubed and shared their content to organize and participate in what they seek, and the government has organized and participated and taken steps in its nature. It shows how critically vital this platform is, that you have created.

That many people will literally fight and die for.

Many ordinary, powerless people are indeed willing to fight and die. But is ICANN going to help them? Or at very least make sure that their decisions won't help those who want to muzzle them?

As Beckstrom and ICANN's board grapple with competing visions of how ICANN should evolve, will the right of ordinary citizens around the world to express themselves - sometimes in ways that companies and governments don't like - be adequately considered? It's hard to tell.

Do ICANN's recent actions and immediate plans reflect the same respect for free expression that the new CEO claims is worth fighting and dying for? I wish I could say "yes" but many people are concerned that the answer is currently leaning more in the direction of "no."

ICANN is about to implement a massive expansion of Internet real estate, most likely starting next year unless the commercial interests opposing the plan succeed in stalling the process. New "general Top Level Domains" (gTLDs) will allow anybody with the resources to apply to run a new "top level domain" registry (the part of your Internet address that goes after the dot, e.g.: .com, .net, .asia, .mobi, etc.). At the same time, ICANN also plans to implement "internationalized domain names" (IDNs). That means it will be possible for people to set up TLD registries in any language or script - including Chinese, Arabic, Farsi, Urdu, Thai, Bangla, you name it. A concrete example: after some people (yet to be determined: more about that later) have set up some IDN gTLD registries in various non-English scripts,  the citizen media community I co-founded, Global Voices, will be able to set up a range of domain names in those scripts for our "Lingua" websites in a range of different languages. In Chinese it might be: 全球之声.组织 ...and so forth.

For the first time, web addresses will be truly multilingual. This represents a huge step forward toward realizing the dream of a truly multicultural, multilingual Internet that is fully accessible to people all around the world. The next several hundred million Internet users on the planet will not be English speakers, and many won't have any education in languages that use anything resembling the Latin alphabet. The next hundred million people who will join the Internet are going to be largely from developing nations, will be increasingly rural, and are going to have much lower levels of education - meaning they're a lot less likely to be the kind of educated elites who are going to feel comfortable typing English-language URL's into browsers.

Let's return to Beckstrom's Iran example. Right now, the Internet is largely the province of Iran's elites. There are a number of economic and political reasons for this. But right now there remain serious linguistic barriers to a rural Iranian with a primary school education seeking information - let alone expressing him or herself - online. With IDN gTLDs, it will for the first time become possible to interact with the Internet 100% in Farsi. Language is not the only barrier to connecting ordinary, less wealthy or educated Iranians living outside Tehran to a global community of Farsi-speakers. But I also think that English-speakers tend to gravely underestimate just how great the language barrier really is to non-college educated people who speak languages not written in Roman scripts.

It's in the interest of global free expression for IDN gTLD's to be implemented as soon as all the basic issues of technical interoperability and basic security can be worked out. However many large trademark owners (mainly companies) are concerned that new gTLD plan will increase their costs and lead to cybersquatting. If new gTLDs are to be implemented, many companies are arguing, there need to be ample safeguards in place for trademark holders. In response ICANN commissioned its intellectual property constituency to formulate a set of recommendations. The Implementation Recommendation Team's final report recommends a number of measures including a global IP Clearing House to mediate trademark disputes, a Globally Protected Marks list and a "Thick Whois" which means that registries will have to retain more consolidated information about domain name owners.  Click here for a critiques and concerns submitted by a range of organizations and individuals. Privacy advocates are concerned about the concentration of data and whether this will increase the chances of registries contributing to more Shi Tao cases. Free speech advocates have long been concerned about the use of intellectual property and trademark claims to stifle legitimate free speech. These concerns abound in spades as regards the IRT report.

In keeping with this post's Iran theme, one of my favorite recent free speech critiques of the IRT is titled Mahmoud Ahmedinejad and the Globally Protected Marks List, by Graham Chynoweth. I quote at length:

At first blush, Mahmoud Ahmadinejad and the Globally Protected Marks List (GPML) do not seem to have anything in common. The first is a politician of debated repute that is seeking to quell disputes over the legitimacy of his election. The second is a recommendation that seeks to protect trademark owners and consumers from an explosion of infringement and source confusion that could be wrought by the introduction of new Top-Level Domains (TLDs). However, upon a closer analysis, they do share one common flaw: both have arguably failed to appropriately prioritize the right to free speech.

In Iran, Ahmadinejad, according to major new organizations, leads a government that has placed restrictions on what the press can cover and publish and what people may say in public. Clearly, some, and certainly Ahmadinejad, would argue that these restrictions on speech are necessary to protect the orderly occurrence of commercial and social intercourse in Iran. However, equally clearly, many would argue that these dictates violate at least one traditional, admittedly American, norm regarding speech—the prohibition on its 'prior restraint.' In order words, it isn't 'fair' or 'right' that the Ahmadinejad government has stopped people from reporting or speaking before it is known whether what they intend to say or publish will actually cause unrest.

In the IRT Final Draft Report, a recommended new mechanism, the GPML, is proposed, which will:

I. prevent individuals from applying to create a new TLD that is identical or confusingly similar to certain 'globally protected marks' until the party interested in running the TLD has gone through a not insignificant administrative process and proven "that the applied-for TLD is not sufficiently similar (visually, aurally, and in commercial impression) as to be likely, as a matter of probability and not mere possibility, to deceive or cause confusion or that it otherwise has legitimate rights to use the applied-for TLD", IRT Final Draft Report, p 19., and

II. prevent individuals from registering domain names in a new TLD that are identical to certain 'globally protected marks' until they prove that the "registration would be consistent with generally accepted trademark laws; namely, that its use of the domain name would not infringe the legal rights of the GPM owner." IRT Final Draft Report, p 19.Or more explicitly stated, in such cases, "[t]o overcome the block, the applicant must show that it has a right or legitimate interest in the initially blocked name." Id.

Just as in the Ahmadinejad case, some would argue that these restrictions are necessary to protect the orderly occurrence of commercial and social intercourse on the Internet. However, also as in the case above, many would argue that these rules violate the traditional American and Internet norm that prohibits 'prior restraints.' In other words, it isn't 'fair' or 'right' for ICANN to stop people from creating TLDs or registering domain names before it is known whether the use of a word in the TLD or domain name, in fact, infringe upon the rights of a trademark holder.

Given the centrality of the idea of 'prior restraint' to this post, it seems important to outline the concept more definitively. Thus, more specifically, 'prior restraint' is the creation of any mechanism that effectively prevents or retards an individual from speaking or publishing material until they can prove the truthfulness or legitimacy of words they intend to speak or the material they intend to publish. See Near v. Minnesota, 283 U.S. 697 (1931). In the Ahmadinejad case we see a 'prior restraint' on speech until the speaker or publisher can show the Iranian government that the speech (broadly construed) will not to cause unrest in the population. In the GPML case, we see a proposed 'prior restraint' on the publishing of a TLD or domain name into the DNS until the publisher can prove that they have a 'legitimate' interest in the word or words at issue.

Later he writes:

Given the importance of the importance of the concept of 'prior restraint' in the context of the American legal tradition, not to mention in the context of the Internet's cultural tradition, it is difficult to imagine how the bestowing the right of 'prior restraint' on the owners of 'globally protected marks' meets this test. More specifically with regard to the latter issue, given the centrality and strength of the prohibition on 'prior restraints' in United States constitutional jurisprudence and that ICANN is inextricably linked and/or given authority by a federal agency of the United States government, it seems apparent that any opponent to the proposed scheme would have, at least at first blush, a credible argument to make before US courts, should they desire to prevent the implementation of the IP Clearing House recommendation into the new TLD space.

Beckstrom made it pretty clear to the New York Times last week that he thinks California Law (which last time I checked was under U.S. law) is good for ICANN and good for the Internet. Will he make that case to the IRT-supporters at ICANN?

ICANN's beleaguered Non-Commercial Users Constituency - the body in ICANN set up to represent non-commercial grassroots interests including those of, say, Iranian student protesters trying to use the Internet to organize and get their message out - has issued two long position papers, one on the substantive problems and one on the procedural problems of the IRT. In a nutshell, they write:

"the substantive IRT recommendations take ICANN far afield of its technical scope and mission, create substantive new trademark rights (beyond existing law), gut existing safeguards and fair procedures for domain name registrants in the Uniform Dispute Resolution Policy (“UDRP”), and create an unbounded situation for abuse by trademark lawyers and those representing trademark owners."

If you're interested in working with the NCUC to advocate for non-commercial grassroots interests in the domain name system, apply to join the constituency here (having no funding and no paid staff, their site is not the most impressive ever to be seen, but they're working to build it with a group of volunteers).

The NCUC itself is fighting for fair representation within ICANN's allegedly bottom-up, grassroots, and inclusive decision-making process. Right now they are under-represented on the GNSO Council (one of the councils that inputs to the iCANN board which then makes final decisions). Efforts to get equal representation for NCUC delegates on the council, in proportion with commercial interests, have met with roadblocks. It's a long, political story. Click here and here for all the gory details. In a nutshell: first, the NCUC is having a charter imposed on them which create a very top-down structure, as opposed to the more bottom-up and democratic structure agreed upon by the community; secondly, the NCUC is only allowed to elect 3 of its 6 council seats: the remaining 3 will be appointed for them by the Board. NCUC chair Robin Gross writes:

Welcome to "bottom-up" policy making at ICANN: where participants are invited to build a "consensus" among a broad range of interests, only to have that consensus discarded by ICANN as a result of relentless insider back-channel lobbying from special interests.

Apparently we noncommercial users wasted our time building consensus among global civil society and participating in a public discussion forum, when we should have been lobbying ICANN board members and ICANN executive staffers -- since that seems to be the only channel of public input ICANN feels accountable to.

Obviously, noncommercial users will never be able to effectively participate in a policy development forum that is predicated on and dominated by insider lobbying from entrenched commercial interests.   ICANN's Board of Directors has a responsibility to the global public interest to ensure noncommercial interests can play a meaningful role in ICANN policy development despite its lack of economic backing. Unfortunately protection for noncommercial interests is systematically being squeezed-out of ICANN's policy development process by commercial interests.

Many people representing commercial interests might argue that simply by engaging in Internet business and connecting people around the world, the needs and interests of students in Iran or ethnic minorities in China are ultimately being met anyway. But experience with Microsoft, Yahoo, Google, and other companies in China has shown that the real world is not so simple. Companies do not always have the human rights interests of their users foremost in mind - understandably, they're seeking to maximize profits for their shareholders. As result, however, free expression can become collateral damage. Google, Yahoo, and Microsoft finally recognized this to be true and signed on to the multi-stakeholder Global Network Initiative, working together with non-governmental free speech organizations, academics, and socially responsible investors to make sure that they conduct their business around the world in ways that do not trample on users' rights to free expression and privacy.

I hope that like those three Internet giants, ICANN will also recognize that even the most well-meaning Internet entrepreneurs from Silicon Valley are capable of making very bad mistakes. Ordinary non-commercial Internet users all around the world who use the Internet to improve their communities and countries would very much like to help ICANN build an Internet that can support and expand global freedom. But most of these people don't have budgets to fly around the world to ICANN meetings, as companies and governments do. ICANN needs to make an effort to ask questions, seek out these people's concerns, and really listen.

Let's hope that Beckstrom will live up to what he said in Sydney. Nobody expects him to die for free speech as did the Iranian activists he praised. But we certainly hope that he will make a genuine effort to support and protect it.

Xinjiang Info-War

The screenshot above comes from the Anti-cnn.com website as it appeared on Saturday. The item is titled "Rebiya Fakes It! Using a fake photo to twist the truth in the Urumqi incident"

Click on the image to view an image of the full web page. I happen to have saved the page (a research habit I've developed) before Anti-CNN.com went completely offline, sometime on Sunday. For the time being at least there is a Google's cache of the front page from July 9.

Exiled Uighur leader Rebiya Kadeer mistakenly made prominent use in interviews of a photo that turned out to be from riots in Shishou, Hubei province, in late June. Al Jazeera has an account here.

Roland Soong at ESWN has a full account of how the photo came to be misconstrued and misused (scroll down to find the relavant material in his long compilation of news about the Xinjiang riots).

Apparently, the source of the error was Reuters, who had sourced the photo from Twitter and put it out on the wire before recalling it. Roland somehow got ahold of Reuters' recall notice (click to enlarge):

The Chinese media and netizens naturally went to town on the whole thing.

The unfortunate - but it appears genuinely honest - mistake by Rebiya Kadeer sparked a fury of comments on anti-cnn, many of which denounced the Western media for emphasizing Uighur casualties while making light of Han Chinese casualties. Some comments on that page and on other threads on the anti-cnn (which unfortunately I did not save) described the July 5th race riot as an act of terrorism, accused the Uighurs of being terrorists, and accused Western governments and the Western media of supporting anti-Chinese terrorism.

Such views were egged on by commentaries in the Chinese state-controlled media, some of which even appeared in English. One, which by Sunday had been removed (but which can still be found in the Google cache) began this way:

By whatever calculations, the blood-thirsty maiming and slaughtering of civilians, as young as six years old, in Urumqi, northwestern China's Xinjiang on July 5, is heinous homicide, barbarity against humanity, and terrorist act on China.

A look into the aftermath of the bloodbath found it bore the hallmark of secret and well choreography aiming at innocent human lives, identical to Al Qaeda's killing of thousands of office workers at the World Trade Center twin towers on September 11, 2001. The terrorists this time did not use flying petrol bombs to detonate tall buildings, they brandished steel rods and wielded knives to end lives.

It continues:

The barbarity has astonished China and the world. Thanks to the authorities' revised policy of free on-spot reportage, in sharp contrast to previous media controls, more people on the globe have got to know the senseless killing. Some said that even during New Stone ages, our ancestors, though barely dressed, did not do this to each other while chasing nuts and edibles in the woods.

..and so on.

This kind of language has raged all over the Chinese Internet between July 5 and Saturday or so.

With Xinjiang we've seen substantial evolution of the Chinese government's media strategy in times of unrest, aimed to make the best of a bad situation. The government has come to recognize that media blackouts don't work in the Internet age. Assuming your goal is to maintain the central government's power and the Communist Party's overall legitimacy (rather than total social control which they gave up on a long time ago) when localized unrest flares up it is more effective a multi-pronged strategy, as follows:

1. Cut off the Internet and mobile messaging in the immediate area where the violence took place.
2. Censor blogs, chatrooms, search engines and social networking sites heavily to prevent people from spreading unofficial information or using social networks to organize. Increase blocks on overseas sites. Shut some domestic ones down if needed.
3. Get Xinhua, the People's Daily, CCTV and other officially sanctioned news outlets on the scene as soon as possible. Fill the airwaves, news pages, and domestic websites with the government-approved version of what happened. (This new approach first emerged as a riot-information-management strategy with last year's Weng'an riots.)
4. Grant access to foreign media - the lesson of last year's Tibet unrest is that if you keep them out they're not going to believe a thing you say. This time, foreign reporters have directly reported about Han victims of Uighur violence as well as Uighur victims of Han violence. Which is probably one reason why the foreign media has not directly challenged the government's official death toll reflecting many more Han dead than Uighur dead, though they've quoted the Uighur exile groups who say the Uighur toll is actually much higher. Coverage instead emphasizes how hard it is to figure out what's going on, which is a much better storyline for the Chinese government than "they won't let us in, what are they hiding?"
5. Be helpful: facilitate the foreign media coverage with press conferences, a dedicated news center, and a discounted hotel in the area you want them to stay in. Allow the police to kick them out of places you'd rather they didn't go to, act coy when reporters complain.
6. Wait for the inevitable mistakes to be made in the Western media - mis-captioned photo here, mistakenly used video there, a grossly oversimplified turn of phrase comparing a race riot to the 1989 democracy movement, plus some downright factual errors here and there. Let the nationalistic blogosphere and Chinese media outlets like Global Times rip into these things as proof of the West's anti-China bias and deliberate obfuscation of the truth.
7. Take advantage of the failure by Western commentators, exiled activists, and human rights groups to consider how their statements and actions may sound and look to ordinary Chinese people, even those who are open to critiques of their government. Last year we saw a pro-Tibet protestor accosting a wheelchair-bound Chinese athlete. In the Xinjiang case, many Western commentators and human rights groups have condemned methods used in the crackdown, expressed legitimate concerns about an impending witchunt against Uighurs, and rightly critiqued the Chinese government's bad policies that led to the ethnic tensions, but then in many cases failed to deplore the killing of innocent Han. This has given many Chinese the impression that the West condones Uighur violence as excusable because they're an oppressed minority. For example, this statement by Human Rights Watch is unlikely to play well with Chinese audiences because it expresses no concern for the loss of innocent Han lives and stops short of condemning all killers regardless of ethnic background. The Chinese government can afford to dismiss it without domestic political cost.

The result is that while they've got much unpleasantness to deal with, their legitimacy in the minds of the majority of Chinese is sufficiently maintained. While many people may have a lot of serious issues with their government, enough people end up concluding that the foreigners and the exiles may not have the Chinese peoples' interests at heart either - so might as well stick with the current crop of bums and work things out with them gradually.

This weekend, however, the censors seem to have decided that the patriotism may have gone too far. Hence the deletion of certain articles, the shutting down of Xinjiang-related discussions in nationalist-leaning fora like the Global Times, and the outage of anti-cnn.com.

One reason for dialing things back may be the reactions coming from the Islamic world. Turkish Prime Minister Erdogan has condemned China's crackdown in Xinjiang as genocide. Global Voices' Iran editor dug up this Persian-language blog comparing the Chinese in Xinjiang to the Israelis in Palestine. I get the impression that there is a lot more of that kind of sentiment out there. China has spent the last few decades cultivating strong relationships with the Islamic world, including Iran. China badly needs their oil, among other things. Chinese companies, engineering crews, and construction workers are all over the Middle East and Africa. Suppress the Tibetans however you like and your external economic relationships won't suffer that much. Treat China's restive Muslim minority in an insensitive, ham-fisted manner, and there could be all kinds of hell to pay. Time to start figuring out how to govern a multi-ethnic, multi-cultural country in a sustainable, enlightened way.

What's also interesting, I find, is that this year in the wake of the Xinjiang riots I've been seeing a lot of discussion and critique on Chinese-language websites about the Chinese government's ethnic policies. Somebody please correct me if I'm wrong but there seems to be a more detailed and nuanced policy discussion going on this year than in the wake of last year's Tibet unrest. A number of postings by Han Chinese people who've lived in Xinjiang, and who think that the government's minority policies have been a failure, have been translated by China Digital Times, Global Voices, Fools Mountain, and ESWN among others. I've come across several Chinese blog posts and articles analyzing the policies on ethnicity and race practiced in Europe and North America. There seems to be a pretty strong consensus among nationalists as well as liberals that - whatever the solution may be - the status quo policies are not working. Combine this with pressure from the Islamic world to do a better job, will things change?

The problem, however, is this: does the Chinese government have the ability to conduct credible policy reform? If good policy happened to be formulated, does the center have enough control over the localities to actually implement it effectively?

If the answer to those two questions continues to be "no," the Chinese government's new and improved information management strategies may help them keep the country together in the short and maybe even medium term, but in the long run even the cleverest and most thorough strategy of censorship and information management will be hard pressed to prop up failed policy and bad governance.

[UPDATED] Green Dam reincarnated in QQ? - initial comments received cast doubt..

UPDATE (Monday 10am London time): Comments received to this post so far indicate that the Boxun.com upon which this post is based may have significantly exaggerated the situation. I am be investigating further and will report back. Clearly I posted prematurely and for that I apologize.

----

Pseudonymous blogger dafengqixi, writing on a Hong Kong-based blogging platform, has linked to this very interesting post published Saturday on the overseas Chinese news platform, Boxun.com (blocked in mainland China).

It reports that QQ, China's largest instant-messaging and micro-blogging service (with over 570 million registered users of its IM service, see Wikipedia entry about it here), is now requiring users of the QQ IM client to install a censorware upgrade - possibly also containing spyware - before they can continue using the client.

It features this screenshot:

Upon opening his (or her - the author is not named) QQ client, the reporter was greeted with the above pop-up window. It says:

QQ Online Upgrade

Your QQ must install the upgrade below before you can continue to use it:

Upgrade content:

Security, Important Patch Upgrade

Detailed information:

Improves the user's login performance

The Boxun article reports that due to recent unrest in Xinjiang and related concerns about security and social stability, the government has required QQ to implement a mandatory software upgrade which would block and track the use of sensitive keywords, enabling authorities to follow conversation trends of users, track and monitor them, etc. The model, according to the reporter, is Green Dam.

The QQ client only works on Microsoft Windows computers. As a Mac user I am unable to download and duplicate the reporter's experience. I would like to hear from others confirming whether they get the same popup requiring the mandatory download. I also hope people will test the upgraded client with some of the Xinjiang-related keywords found in a recent government order to filter search engine results, and some of the political keywords recently unearthed from the Green Dam software.

For readers unfamiliar with QQ, here is how Scott Harley at the Internet and Democracy Blog recently described QQ's importance Chinese Internet users:

Another emerging form of domestic communication is Tencent’s Instant Messaging (IM) on QQ.com. No site in China enables greater horizontal web communication than QQ.com, now the 9th largest web property in the world. Founded by Pony Ma in 1998, Tencent –a Chinese-listed company earning $1.2 billion annually in revenue, 88 percent via the sale of “virtual goods” rather than online advertising– has over 570 million registered users of its IM service. In January Tencent launched an English version of the IM platform at IMQQ.com, and a 3G version that offers QQ chat, real-time news, and search engine accessible over mobile phone.

Many users in Western China, and across rural China, do not have email accounts. And many rural Chinese view and understand the Internet as Tencent, the platform on which they’ve grown up. In fact, as of March 2009 China had as many active Tencent QQ users as it officially had people online. Despite focus on access to Google and Twitter –observations of Tweet trends, and Google search engine query data patterns– undoubtedly most relevant in China is continued access to those domestic services of communication most widely used by Chinese citizens.

China @ ICANN: thoughts from former CEO Paul Twomey

Nobody else appears to have reported this - at least not anywhere I can find - but last week marked a major turning point for China's engagement with ICANN. It was probably also a major turning point in China's strategy on Internet governance.

The Chinese government sent Cui Shutian, Deputy Director, Division of Telecommunications Services and Resources, Ministry of Industry and Information Technology (MIIT) to represent its interests in ICANN's Governmental Advisory Committee (GAC) at last month's ICANN meeting in Sydney. Outgoing ICANN CEO Paul Twomey confirmed to me that it was the first time the Chinese government has engaged directly with ICANN since 2001.

(It's likely that some of my more China-focused and less-techie readers have never heard of ICANN - the Internet Corporation for Assigned Names and Numbers. This California-based non-profit corporation is responsible for making sure that the Internet functions as one globally inter-operable system. Its primary job is to coordinate the global assignment of domain names and IP addresses - which turns out to be a very complicated and increasingly political job. Click here, here, here, and here for useful background.)

One reason the Chinese government disengaged from ICANN and is now re-engaging will be familiar to China wonks: Taiwan. The issue of what it should be called. Beijing was not interested in lending any legitimacy to Taiwan's government under the pro-independence president Chen Shui-bian, and the Chen government wasn't big on compromising with Beijing either. Now it's agreed that Taiwan will officially be referred to as "Chinese Taipei" at ICANN (though the .tw designation won't change), and the two governments in the Ma Ying-jeou era are more willing to be pragmatic with one another, in cyberspace as well as in "meatspace."

The second reason for China's re-engagement with ICANN is that the Chinese government and ICANN have both realized they need each other, at least for the short and medium term.

Some background (skip this paragraph if you are an Internet governance wonk): The Chinese government has long suported Internet governance reform. The fact that ICANN (founded in 1998) ultimately answers to the U.S. Department of Commerce has for most of this decade been a matter of concern to a number of governments, from China to Brazil, Iran, and the European Union. Reform proposals have ranged from ICANN "internationalization" - with supervision by multiple governments - to scrapping ICANN completely and transferring its functions to a U.N. body like the ITU (International Telecommunication Union). Arguments over ICANN's future reached a climax in 2005 during the run-up to the World Summit on the Information Society in Tunis, at which governments agreed there wasn't enough consensus to change the ICANN-based status quo. Instead, the Internet Governance Forum was created with a 5-year mandate to continue inter-governmental and multi-stakeholder dialogue on how the global Internet should be governed, managed, developed, and regulated in the future. That mandate will end next year. In May of this year China publicly opposed the renewal of the IGF's mandate, declaring it a costly, messy, time-wasting shop and reiterating its longstanding position that Internet governance should be in the hands of sovereign governments, not other groups. Meanwhile the Joint Project Agreement between ICANN and the U.S. Department of Commerce expires this September. The U.S. Congress, concerned with U.S. interests, wants to keep things as they are. The European Union issued an official statement last month calling for "a more open, independent, and accountable governance of the Internet." African nations, interestingly, have shifted from supporting reform to supporting the status quo. Chinese officials have continued to express concern about a "monopoly" controlling the Internet, and have made it clear that they want to continue discussing the JPA, but I've not seen anything to indicate an official Chinese comment on the EU position or any other government's position.

According to Paul Twomey who stepped down as CEO of ICANN last week, China has not recently made public statements on ICANN and the JPA. Meanwhile he, the ICANN board, and the other members of the Governmental Advisory Committee have been working hard to make China feel comfortable engaging with ICANN. Here's what he said to me in response to my question about China's position:

In terms of the relationship with the United States government and the Joint Project Agreement, I haven't actually heard what the position of the Chinese government is but they have publicly said things in the past. But I dont think we're in anything like the bipolar situation that we had three four five six years ago, where we're sort of "ICANN love it or hate it." I don't think it's in that space anymore. I think it is, much more pragmatically, that this is an institution where there's a space for the Chinese government to participate in, that it's looking after the interests of the Chinese internet community, that they're dealing with real issues that really affect their concerns, and they're welcome. And i think that's important.

China can't afford not to engage with ICANN at this point in time. The Internet is about to undergo a huge real estate expansion and the Chinese government - along with China's domain name registrars - wants to make sure that Chinese government interests are well served as the rules and technical arrangements get laid out between now and early next year.

In 2010 ICANN will implement two big changes, and China has a big interest in how these changes are implemented. First, ICANN will soon allow anybody (who can pay the six-figure registration fee) to apply to run a "generic top-level domain" (gTLD) (Explanation for non-wonks: .com, .net, .cn, .asia, .mobi, .org, .gov, et cetera are all "top-level domains;" the word before the dot, for example "cnn" in cnn.com is called a second-level domain and that's what individuals, organizations, and companies buy when we purchase a domain name for our website. So for instance, if I was extremely rich and had the technical resources I could apply to create and operate .rebecca). So a religious organization, a political party, a company, or anybody with the resources who wants to administer a distinctive Internet address can apply to establish a new gTLD.

But that's not all. In 2010 ICANN will not only allow more gTLD's to be created, but it will also enable the creation of "internationalized" top-level domains, in non-English/non-Roman letter scripts. In other words it will for the first time be possible to have internationally accessible top-level Internet addresses in Chinese, Arabic, Farsi, Urdu, Thai, Japanese, and whatever other language can be input onto a computer. This is huge because it will make the Internet much more accessible to non-English speakers who have difficulty dealing with the current English-based global domain system. If we really want a truly global and multilingual Internet, having "international domain names," or IDN's as ICANN calls them, is essential. These new international top-level domains (IDN TLDs) will be divided into two categories: "country code TLDs" (ccTLDs) and "general TLDs" (gTLDs). For existing English-character ccTLD's (like .cn) each relevant country gets jurisdiction over how it is administered and the same will be the case for international ccTLD's. So .中国 will be controlled and administered the same way as .cn, and websites under that top-level domain will be subject to Chinese law. (Chinese bloggers have informed me that they stay away from .cn domain names because they lose their web address if their website is too politically controversial.) gTLD's, however, are different. ICANN's intention is that anybody anywhere in the world (with adequate resources) can apply to run a Chinese language gTLD. I asked Twomey what happens if people in, say, Canada or Australia apply to run .falungong and the Chinese-language equivalent. Here's what he said in response:

First of all our process, and the process of moving our policy forward, are neutral processes. And we have a series of objections mechanisms through which people can bring objections, one which we're still working through, which is morality or public order, which is a term that exists in international treaties. There's still a great deal of discussion including in the Government Advisory Committee about how that can possibly work, and should it even be there.... But ICANN is not in the business of the application level. We're not in the business of content. So the strings that people might put forward, were not in the buisiness of deciding whether its a good string or not. There are opportunities, but we're not in the business of saying that's a good string or a bad string. We dont like that one we like this one. And its global. The generic top level domains are global top level domains.

So, in other words, I followed up, anybody can apply to run a generic TLD in any language from anywhere? He replied:

That's right. Its a global technology. The technology doesn't recognize geographic boundaries so we support that. Now whatever governments might decide to do in terms of access or filtering is their business not ours... and we leave that.. because we are if you like the guardians of the single inter-operable internet, our community and the ISPs [internet service providers] are the people who provide the single inter-operable internet. We think it's very important that the issues of the addressing and routing system are separate from the issues of content carried on them. And so we don't comment, we dont condone, but we don't make comment upon those sort of content issues. But what we do say is: you know, if a government has a content issue, don't break the domain name system to try to fix it. Because that's like killing the goose that laid the golden egg.

Finally, here's what Twomey said about jurisdiction of gTLDs:

Let's make it clear that generic top level domains, when they're created, will have contracts with ICANN. And those contracts will clearly state that the applicable law is the law of California. And they will be contracts. Right? You know, any national laws that apply to end registrations we can't comment on, but the law under which the TLD will actually operate will be the law of California.

Many of the details, however, in terms of how a government or anybody else can object to the creation of a new TLD on the grounds of "morality and public order," and under what criteria the ICANN board then comes to a final decision, have yet to be worked out. Also, in the case of multiple applicants for the same TLD, there will be some kind of process for the ICANN "community" to decide who deserves to get that particular domain. So being there at ICANN meetings is very important if you want to influence how the rules get shaped and who has rightful claim over various names. Having the ear of board members by developing a personal relationship with them is also very important, I'm told by people who currently run generic TLDs.

Then there are trademark and other issues related to who has the "right" to a particular TLD name. There is already a big fight over the extent to which brand names can be protected or reserved by companies that are worried about their trademarks being (in their view) appropriated by others. The Chinese government and Chinese companies have an interest in trademark protection not overly favoring Western business and industry on the one hand, while still protecting Chinese companies on the other, while also making sure that arbitration mechanisms are properly internationalized and not overly Western-centric. There are also concerns about who has a right to register and administer, say, .beijing or .guilin - to name a couple examples. The Chinese are also concerned that fights over trademark protection and proposed attempts by ICANN to create a new trademark arbitration body will delay the roll-out of gTLDs in general, and internationalized gTLD's in particular.

Then there are a bunch of technical issues related to this multi-lingual rollout, including an appeal by the Chinese, Japanese, and Koreans among others that ICANN must change it's current rule that non-country TLD's must be 3 characters or more - given that most Chinese words are two characters in length, the three-character minimum requirement is obviously unreasonable. While the Chinese have actually already implemented Chinese-language TLD's within mainland China, they don't work outside of China. Asian ICANN participants last week repeatedly expressed frustration that it's taking too long for ICANN to move forward on this.

So there are lots of reasons why the Chinese need to engage with ICANN now, and not wait till it's restructured or its functions re-appropriated to its greater liking. Here's how Twomey put it:

First of all there's a much better understanding of what ICANN does, what it is and what it's not. I think secondly ICANN is working on issues that really are important to the internet communities of China, for instance. So particularly on the internationalized domain names and the internationalized domain names for the country codes, these are core areas of interest to the Chinese, especially to the Ministry of Information Industry and Technology. But also for the broader Chinese internet community. I'm very pleased that not only has China come back to the GAC but that there's a broad commitment in China now to have Chinese institutions and organizations participating in the ICANN context. ...ICANN is looking at the role of the GAC and what it does and how it participates. And how it interacts with the board. we've continued to have that discussion. I think it's useful that China's at the table to have that discussion.

I was a first-timer at ICANN last week, but sitting through a week of meetings, I got the distinct impression that few people expect ICANN to be radically restructured this year. Most people seemed to expect that the JPA will be renewed in some fashion for now, due to lack of consensus over an alternative. Meanwhile the Government Advisory Committee seems to be getting an expanded role, with more opportunity to lobby the interests of governments as ICANN moves forward in radically expanding the Internet's real estate. How things evolve in the longer run is harder to say, but the people running ICANN are clearly trying to make governments feel welcome to engage and work for change from within - pointing out that this is more likely to be conducive to a stable global Internet - rather than try to dismantle or radically and suddenly restructure it from the outside.

Bringing in China was likely a critical step by ICANN to ensure its own survival, at least in the short to medium term. For now, ICANN and China need each other.

Green Dam is breached...Now what?

As a number of China hands predicted, the Chinese government has postponed its mandate requiring that all computers sold in China must include the Green Dam -Youth Escort censorware by today.

Yesterday after the news broke I told the Financial Times: "There's been this impression in the internet industry that when the Chinese government makes a demand, they have to roll over and play dead. The lesson here is that's not necessarily the case." I'll put it more strongly here: The Green Dam episode proves yet again that when companies respond to critics by saying things like: "it's beyond our control if we want to do business in China" or "there's nothing we can do or we will get kicked out," that is a huge pile of, well, equine excrement.

What should be abundantly clear from events of the past several weeks is that industry and even ordinary citizens can have a real impact on policy outcomes in China - especially information and technology policy that has a direct impact on large numbers of people nationwide. It's not clear whether a new deadline will be set, the mandate radically revised, or the whole thing quietly scrapped. Which of those three options becomes reality depends very much on the actions over the next few months by industry, Chinese netizens, and various other actors.

Andrew Lih writes:

This should be seen as a case study on how the complexities of China’s decision system is much more nuanced than what a “Communist” regime would suggest, and the role of citizen deliberation in a new, upwardly mobile, aspirational, IT-savvy China.

While the outside world sees the PRC government in absolute control, in reality the heavy handed, top down authoritarian system rides on a delicate balance of, bottom up public consent that supports the state’s legitimacy.

Sky Canaves at WSJ.com's China Journal has a great post titled Green Dam and the Politics of Consent, in which she points out:

While it’s impossible to know what impact Chinese critics of the plan had relative to the vocal protestations of foreign governments, trade organizations and PC makers, the ministry’s latest move on Green Dam highlights the consensual nature of government and politics in China, which emphasizes stability and agreement among all parties involved.

Dan Harris sees this as an example of how China really is moving in the right direction:

China is billing it as a delay, but I can virtually guarantee this software will never be heard from again. I say this for two reasons. One, the people did not like it and Beijing does NOT want to go against the people on something like this. Since there is absolutely no reason to believe the people will ever start liking something like this, there is absolutely no reason to believe the software will return. Two, I know movement has been slow, and I know it has been in fits and starts, but if we were to draw a straight line through the rises and falls, freedom is on a fairly inexorable march in China.

William Moss, a.k.a. "Imagethief," however, isn't ready to pop the champagne just yet:

Green Dam Youth Escort may have submarined as predicted, but make no mistake, this isn't over. The objective that drove it the Green Dam plan, the desire to "purify the Internet", still stands. All attempts are being made to wrap the demise of Green Dam in face-saving balm, but there is no disguising the thoroughly humiliating nature of the episode. Nothing stokes the fires of zealotry like a bout of punishing humiliation, and I'd expect to see that zealotry redirected in coming months in an attempt to justify the Green Dam initiative. The recent assault on Google may be a sign of things to come. PC makers may be able to breathe a sigh of relief now, but portals, video sharing sites, social networks, Internet cafes and others may be in for an interesting few months. Watch this space.

A piece in today's Financial Times (subscription required) points out that scrapped mandates sometimes get resurrected later on in somewhat different forms. One example was Beijing's failed attempt to require that all mobile phones sold in China adhere to a new WAPI standard (alternative to Wifi). Strong industry pushback caused that mandate to be scrapped but it's now been revived, the FT writes:

In 2004, the Chinese government surprised the communication industry with an edict that the homegrown wireless encryption standard – in competition to WiFi – would be compulsory for all mobile data products sold in the country. Only just before the May deadline, Wu Yi, then vice-premier, backed down during a visit to the US. After China failed to win recognition for WAPI as an international standard two years later, the industry assumed the episode was over.

But now, WAPI is back, with a demand by the Chinese government that WiFi handsets for sale in the country will only receive approval if they are also equipped with WAPI.

I agree with "Imagethief" that while the Green Dam mandate may never be revived in its present form, we can absolutely expect that Chinese government efforts to clean up Internet smut, promote a "green Internet," and control Internet content in the name of protecting China's children are by no means over.

There is the short run and the long run. In the short run, everybody's unavoidably in for a rough patch which is likely to continue in one way or another through the end of the year with the 60th anniversary of the PRC's founding on October 1st.

In the long run the Chinese public will value companies that treat adults like adults and give them choice and control over their lives. In the long run such products will win consumer loyalty. Ethical business practices that demonstrate respect for users' and customers' interests and rights - despite a very difficult regulatory environment - will serve companies better in the long run. A roll-over-and-play-dead policy is no way to build a brand's reputation in China or anywhere else.

Industry should take the Chinese government at its word that its goal is to protect children, and try to work out sustainable, ethical solutions on that basis. The drive to censor porn does actually have roots in genuine public frustration about children's easy access to smut and violence on the Chinese Internet. There is a real problem and lots of people expect their government to do something about it. Chinese parents' frustration is shared by parents around the world. Whether the problem can actually be solved by technical means without over-censoring and violating civil liberties, however, or whether the problem's real solutions are social and cultural (requiring parents and teachers to do the hard work as parents and teachers instead of foisting most of the responsibility on companies and governments), is subject of debate in China as well as in pretty much every country with substantial Internet penetration.

Because the Chinese government is un-transparent and unaccountable it will keep using porn as an excuse not only to exercise political control but also to go after foreign companies like Google when the mood suits them (with Chinese competitors egging them on) - and maybe even to help give a leg up to home-grown businesses. But it's also important to recognize that the Chinese government views itself as being part of a global pro-censorship, anti-porn freak-out "we have to do something" bandwagon.

Even democratically elected governments are increasingly turning to national-level Internet censorship as a solution to child porn and hate speech, and in some cases also intellectual property theft, and other real or perceived social ills - depending on whose opinion you go by. The German parliament passed a bill two weeks ago to implement a national Internet filtering system. As I wrote in my op-ed for WSJ Asia last month (links added):

In a petition against the bill, German civil liberties groups call it "untransparent and uncontrollable, since the 'block lists' cannot be inspected, nor are the criteria for putting a Web site on the list properly defined." These concerns aren't unfounded: Some German politicians have already suggested extending the block list to Islamist Web sites, video games and gambling Web sites, while book publishers have suggested it would also be nice to block file-sharing sites too.

Since 2007 Australia's Labor government has advocated a policy of mandatory national filtering. In the face of fierce public criticism the censorship plan may be downgraded to a voluntary industry initiative. But critics remain concerned the block list will not be selected and maintained in a transparent or accountable way -- and that the process for appeal is very unclear, making it likely that some Web sites will be blocked in error or that "mission creep" could take place without adequate public supervision.

In Britain, a "block list" of harmful Web sites, used by all the major Internet Service Providers, is maintained by a private foundation with little transparency and no judicial or government oversight of the list. At least one British family protection group, Mediamarch, has already spoken out in support of the Green Dam concept of moving censorship from the network down to the device level.

PC makers, mobile phone companies, search engines like Google and Internet content providers will remain under strong government pressure in China to "do something" about things the government doesn't like - just as today they are under pressure in many other countries from Thailand to Turkey.

Industry should give the Chinese government as little excuse as possible to use child protection as an excuse to accomplish other goals that have much less public support and which are contrary to globally recognized human rights norms. Industry should perhaps encourage and maybe even fund in China a set of public forums and independent research efforts and so forth to examine how can industry work together with China's parents, teachers, and government to protect China's children. Initiate efforts to work with Chinese experts to develop strong culturally appropriate Chinese-language parental control software that puts control in the hands of end users. China is a potential R&D test-bed to innovate on genuine best practices in child-protection technologies... to the extent that we can realistically expect technology to be able protect children from humans...

Ultimately the problem extends well beyond China and the solutions must also be approached in a global way. It's too easy once technical censorship systems are put in place for them to be abused, or for there to be "mission creep" even in democratic societies. There is little consensus in democratic countries on questions like: What are the best ways to protect children in the Internet age? Is too much emphasis being placed on technical solutions? Do politicians and civil society need to recognize that ultimately the solutions are social and cultural? How best to develop child protection policy and business practices that really accomplishes the goal of protecting children without encroaching on civil liberties and giving governments the excuse to censor content that goes beyond porn without appropriate accountability mechanisms? We need a more intelligent global policy and industry discourse on these questions, because as it stands global political trends are not on the side of free speech protections... whatever the trend-line might look like in China on any given day.

(Note: Some content at the end of this post appeared on this blog briefly last week when an unfinished draft post was mistakenly published.)