About GreatFire

GreatFire

GreatFire (GreatFire.org) is a website and organisation, launched in 2011 by an anonymous trio, that monitors internet censorship in China and helps Chinese internet users circumvent it. Its stated mission is 'to bring transparency to the Great Firewall of China' — the system of blocking and filtering, formally the Golden Shield Project, by which the Chinese state controls what its internet users can reach. GreatFire lets any visitor test, in real time, whether a website is accessible from within China; it maintains a large database of blocked URLs, keywords and searches, periodically testing a collection that already exceeded 100,000 URLs by 2015; and it has pioneered 'collateral freedom' — mirroring blocked news sites on major content-delivery networks such as Amazon CloudFront and Cloudflare, so that censoring them would inflict unacceptable economic collateral damage. It built FreeWeibo, an uncensored mirror of censored Sina Weibo content, and later tools including FreeBrowser, AppleCensorship.com and the GreatFire AppMaker. The team has described itself as a small group working from inside China — Beijing is recorded as its nominal city in this directory — whose members remain anonymous and know little about one another, so that no single arrest can bring the project down. Funded by sources inside and outside China including the US-backed Open Technology Fund, it is regularly cited by Reuters, the Guardian, the Wall Street Journal and Bloomberg, survived a state-level counter-attack by the 'Great Cannon' in 2015, and won the 2016 Index on Censorship Freedom of Expression Award for digital activism.

Founding and anonymity

GreatFire was set up in 2011 by three anonymous individuals, tired — in TechCrunch's account — of China's approach to the internet. Its public head uses the pseudonym Charlie Smith, who has said, only half joking, 'I hope we put ourselves out of business.' The anonymity is structural, not theatrical: members remain unknown to the public and, for security, do not know much about each other, precisely so that the detention of any one of them could not compromise the whole project — a precaution whose logic the Chinese state's treatment of other digital-rights and legal-aid workers in the same decade made concrete. The organisation describes itself as working from inside China, which is both its authority — its measurements are taken from within the firewall it studies — and its risk. It began by collecting data about which sites were blocked in China; within a few years it was monitoring thousands of sites, domains and Google searches through a network of computers inside and outside China, testing whether content available on the open internet is reachable within China, showing how much of the time each tested page has been blocked, and explaining the mechanism of blocking where it can be determined. In 2015 the Associated Press reported its funding as coming from a variety of sources including the Open Technology Fund, the US government-backed programme that supports internet-freedom tools; the project's own accounts also acknowledge supporters inside China. That mixed funding and in-country operation have made GreatFire a standing exhibit in arguments about foreign involvement in China's internet — and have never, in more than a decade, stopped it publishing.

The Analyzer: making the Firewall visible

GreatFire's first and most durable product is transparency itself. Its testing system — the GreatFire Analyzer — allows anyone to enter a URL and see its accessibility history from within China: whether it is blocked, when blocking began, and with what consistency. The site also provides Blocky, a system for searching online services and checking their status. By 2015 the database covered over 100,000 URLs, tested periodically to produce a history of availability and restriction for each one; Index on Censorship's account describes the team monitoring thousands of sites, domains and searches at any time. The scholarly and journalistic value of that accumulation is difficult to overstate. Censorship is normally deniable: a site fails to load and no authority announces why. A longitudinal, timestamped public record converts anecdotes into evidence — this domain was blocked on this date, by this method, following this event — and GreatFire's records became a standard citation for reporters and researchers documenting the blocking of Instagram, of messaging apps, of Google services, and of attacks on services such as Apple's iCloud and Microsoft Outlook in China. The organisation's blog and its @GreatFireChina social-media presence report new instances of restriction as they are detected. Its mission statement — bringing transparency to the Great Firewall — understates the ambition slightly: the database is an attempt to make a secretive, shifting system of control into a public, measurable object about which arguments can be settled with data.

Collateral freedom and the BBC

Measurement led to circumvention. GreatFire's signature strategy, 'collateral freedom', rests on a simple piece of political economy: China's censors can block a dissident website at no cost, but they cannot lightly block the cloud platforms — Amazon Web Services, Microsoft Azure, GitHub, Cloudflare — on which Chinese commerce itself depends. By hosting mirror copies of blocked sites on those platforms' content-delivery networks, a blocked publisher becomes, in effect, collateral the censor cannot afford to destroy: 'Collateral Freedom ties access to information to the Chinese economy', as Charlie Smith put it; 'if authorities truly want to block access to this information, then they must give up certain access to economic freedoms.' In November 2014 the strategy reached its widest audience when the BBC worked with GreatFire to make its Chinese-language website — blocked in China almost since launch — available to users there through mirrored delivery via Amazon CloudFront and Microsoft Azure, with no special tools required of readers. GreatFire built similar mirrors for other blocked sites, among them Google and the New York Times' Chinese edition, and for Chinese-language outlets such as Boxun and China Digital Times, maintaining a directory of links on GitHub; with Reporters Without Borders it later joined an operation to unblock nine websites across 11 countries, extending the model beyond China. In February 2015 it released an Android app applying the same approach, allowing users in China and other censored countries to reach the mirrored sites. By its own account in later years, GreatFire has directed Chinese internet users to censored news about corruption, politics and scandal more than 13 million times.

The Great Cannon: retaliation, 2015

The strategy's success provoked the most dramatic counter-attack in the project's history, which became a landmark in internet history in its own right. In March-April 2015, GitHub — then the world's central open-source hosting platform, and host to two GreatFire projects enabling access to GreatFire's site and the Chinese New York Times — suffered the largest distributed denial-of-service attack in its history. Researchers at Citizen Lab, the University of Toronto, Berkeley, Princeton and the International Computer Science Institute established that the attack was powered by a new Chinese system they named the 'Great Cannon': a man-in-the-middle tool co-located with the Great Firewall but operationally distinct, which intercepted foreign users' requests to Baidu's analytics and advertising infrastructure, silently replaced the requested script in about 1.75 percent of cases with malicious JavaScript, and thereby conscripted bystanders' browsers worldwide into flooding GitHub. Baidu denied involvement. Citizen Lab's conclusion — that a state-level actor had moved from blocking content at China's borders to weaponising traffic passing through China against targets abroad — made the Great Cannon a reference case in debates about offensive state cyber operations, compared by researchers to the NSA's QUANTUM system. For GreatFire, the episode was simultaneously validation and warning: its GitHub pages were important enough to draw a weapon of that novelty, and its own site was a named target of the traffic flood. GitHub mostly stayed online; so did GreatFire; and the Cannon has never again been observed in comparably open use — an outcome the organisation and its defenders read as the collateral-freedom logic working at the level of statecraft.

FreeWeibo and the censorship of China's own platforms

GreatFire's work is not only about foreign websites blocked at the border. Its sister site FreeWeibo monitors Sina Weibo — China's vast microblogging platform — and republishes leading content that has been censored and deleted by the authorities and the platform, restoring to public view posts that China's internal, keyword-and-review censorship system removed. The distinction matters analytically: the Great Firewall governs traffic between China and the outside world, but the larger part of Chinese information control happens inside, on domestic platforms, through deletion, invisibility and account action. FreeWeibo documents that second system with the same method as the Analyzer — collect, timestamp, republish — creating a record of what could not be said even by Chinese users to Chinese users. The organisation later extended the same idea to other censored environments, listing FreeWeChat among its products, and built AppleCensorship.com, with Open Technology Fund support, to track Apple's removal of apps from its App Stores around the world, including Hong Kong — an acknowledgment that by the late 2010s some of the most consequential censorship decisions affecting Chinese users were being taken by American platform companies, and deserved the same transparency treatment as the state's. Together, Analyzer, Blocky, FreeWeibo and AppleCensorship define the organisation's documentary half: a set of instruments pointed at every layer of the stack where speech to and from China can be switched off.

Apps, AppMaker and the later toolkit

The circumvention half also kept developing. FreeBrowser, the organisation's censorship-evading browser, gave users in China access to censored stories through a single app; GreatFire has said the experience inspired its most ambitious attempt to generalise the model, the GreatFire AppMaker. Announced in the mid-2020s, AppMaker lets any organisation, media outlet or individual blogger whose site is blocked in mainland China compile an Android app that carries their content past the restrictions — naming the app, supplying the homepage from which it draws, and letting the tool embed GreatFire's circumvention browser, which evades blocks using multiple strategies including machine learning. The Human Rights Foundation was an early adopter, making its content installable in China although its website remains inaccessible there. GreatFire emphasises that AppMaker and the apps it builds are not limited to China: they can be used wherever similar internet restrictions exist — the same internationalisation the Reporters Without Borders operation signalled a decade earlier. The trajectory across the toolkit is consistent: from measuring blocks, to mirroring a handful of important sites by hand, to productising circumvention so that blocked publishers can deploy it themselves — a small anonymous team steadily converting its own hard-won tradecraft into infrastructure for others.

Recognition and standing

GreatFire's standing in its field was formalised in 2016, when it won the Index on Censorship Freedom of Expression Award for digital activism. Index on Censorship's citation described the essentials as this profile has: founded in 2011 by three anonymous individuals to counter the systematic blocking of websites touching on news, human rights, democracy or religion; grown from data collection into monitoring at scale; and distinguished by the collateral-freedom work with organisations such as the BBC and Reuters that pushed uncensored material into China at a scale of millions of readers. Open Technology Fund's Dan Meredith, one of its backers, described its network of in-country and external testing machines and its directory of unblocked material as a resource without real parallel. The press citation record matches the award: Reuters, the Guardian, the Wall Street Journal, Bloomberg and TechCrunch have all treated GreatFire as, in TechCrunch's words, 'perhaps the most trusted authority on the subject' of Chinese internet censorship — trusted, notably, by competing news organisations that send their readers to its data. Fifteen years after its founding, the project remains what it was designed to be: pseudonymous, small, technically inventive, and cited everywhere its subject is seriously discussed.

Assessment: what GreatFire is, and its limits

GreatFire is classified in this directory as an online news publication, and its blog does publish news and analysis about censorship; but it is more precisely a digital-rights organisation whose primary products are data and tools, and whose 'news' is the by-product of measurement. That distinction guides proper use: cite GreatFire as the authority on whether and when something was blocked in China, and on how circumvention works; read its commentary as expert advocacy from a declared opponent of the censorship system, funded in part by a US government-backed fund — a fact its critics in Beijing emphasise and its own accounts do not hide. Its anonymity, necessary as it is, means its internal claims cannot be independently audited in the way a named institution's can; its estimates and self-reported reach figures should be attributed to it. Its strategic premise — collateral freedom — also ages with the architecture of the internet: as Chinese platforms and clouds have grown and foreign cloud dependence has changed, the economic hostages available to the strategy have shifted, which is one reason the organisation's own toolkit has moved toward apps and machine-learning evasion. What has not changed is the founding wager of 2011: that a censorship system which depends on opacity can be fought, measurably and cumulatively, by three anonymous people with testing machines and a database — and that transparency, once published, is very hard to block.