Filling Communist China’s big data lake
Aussie think tank reckons Chinese info intake from open systems abroad equivalent to 20b Facebook photos
Data is the new oil in the sense that whoever possesses it has lots of power, just like with the sticky black stuff. Getting your hands on enormous amounts of data is relatively quick and easy, says a new policy brief, “Engineering Global Consent: The Chinese Communist Party’s datadriven power expansion”, from the Australian Strategic Policy Institute (ASPI) defence think tank.
Rather than coercement and hacks, vast amounts of data is obtained via common tech products that make governance more efficient and online services that people find useful in their everyday lives.
Convenience is king and it helps Global Tone Communications (GTCom), a company that ASPI says is under direct control of China’s Central Propaganda Department, slurp up two to three petabytes annually of unstructured data.
Computer hard disks are typically terabyte sized, and a petabyte is a thousand times bigger. ASPI reckons GTCom’s data ingestion equates to 20 billion Facebook photos.
That enormous amount of data is fed into artificial intelligence systems to build tools to monitor public sentiment worldwide and to digitally manipulate it, along the lines of what Russia is said to be doing.
GTCom collects data in more than 65 languages for processing and translation, for government agencies and corporate clients to use in applications such as facial and voice recognition (perhaps to track down online likenesses of Winnie the Pooh,
who is banned in China).
Local academics are being tapped for the efforts, as “GTCom will work with the International Federation of Translators, the Australian Institute of Interpreters and Translators, the New Zealand Society of Translators and Interpreters, the University of Sydney, the University of Queensland and the University of Auckland to explore deeper cooperation in machine translation and
artificial intelligence and thereby further build a globalised presence of AI language technology. We certainly look forward to it”.
Moreover, ASPI says, GTCom c-ooperates with Chinese tech giants like Huawei and Alibaba Cloud and Haiyun Data, a public security decision-making company, to “embed technology into its products as a means to collect data that goes directly back to servers that the partystate controls, for users that would serve its intent in addition to GTCom’s translation services”.
Maybe it shouldn’t be a surprise to see GTCom chief executive Eric Yu dropped in on Facebook in 2014, and had his picture taken with the social network’s vice-president of corporate development, Vaughan Smith.
ASPI’s report should be read in the context of the current tensions between Xi Jinping’s illiberal and expansionist China, and Western nations struggling to contain a rival that’s feeding data gathered from openness and transparency into a giant control and surveillance machine with global reach.
The release of the policy brief coincides with the United States’ blacklisting of several Chinese AI and surveillance tech companies for human rights breaches against Uyghurs in Xinjiang, and the document again points to Huawei being banned from 5G networks for its opaque CCP connections.
ASPI veers towards propaganda itself when it states that Cambridge Analytica “demonstrated the power of digital technologies when it used the data of 50 million Facebook users to influence the 2016 US presidential elections — and allegedly, the Brexit referendum too”.
That’s a dubious claim at best, and you would have to assume that the huge Chinese data collection effort produces better results than what Cambridge Analytica achieved, or it wouldn’t be worthwhile.
Nevertheless, you can’t disagree with ASPI when it recommends that governments strengthen data privacy laws, and understand how information on us is collected.
Not everyone with access to seemingly innocuous data has the best of intentions, and we really need to know who is trying to digitally influence us.
Not everyone with access to seemingly innocuous data has the best of intentions.