In my chapter for this year’s China Story Yearbook, ‘AI Dreams and Authoritarian Nightmares’, I explored China’s use of artificial intelligence for public security. Writing that chapter less than six months ago, I thought that a major, nationwide crisis would demonstrate how far the Chinese Communist Party would be willing to go to deploy these kinds of technologies for command and control. What I didn’t anticipate was that the crisis would come so soon or would emerge in the form of an invisible, viral adversary. While China’s technological response to COVID-19 has been undeniably innovative, its approach is unlikely to be adopted by other countries concerned about privacy and potential uses of the data that’s collected.
Using technology to fight coronavirus
Despite a halting start, China has used all the tools in its technological arsenal to respond to the COVID-19 pandemic. Artificial intelligence has been no exception. AI models have been developed to predict outbreaks, analyse patients’ CAT scans, offer online medical consultations and speed up the sequencing of the coronavirus genome. Meanwhile, robots and drones have been deployed to sterilise public spaces, enforce physical distancing, take temperature checks and deliver food and medicines. Other countries are now taking similar measures in the fight against COVID-19.
Few countries, however, have gone as far as China in using technology to enforce quarantine and lockdown measures. In a February episode of The Little Red Podcast, Orville Schell from the Asia Society argued that even before the pandemic began, Xi Jinping had espoused a form of “techno-authoritarianism” that would have surpassed George Orwell’s worst nightmares. But COVID-19 has provided the logical pretext to ramp up social controls and surveillance even further.
Take for example the Alipay Health Code. Launched in Hangzhou on 11 February, the app reached 50 million Chinese in over 100 cities within a fortnight. The app uses machine learning to classify a user’s health status and assigns them a QR code that is either red, yellow, or green. Red means two weeks of quarantine. Yellow means self-isolation for one week. Those with the coveted green code can move freely and access public spaces, provided they scan their code at regular checkpoints.
The health code system is dynamic — changing in real-time based on the user’s movements, self-reported health status, changes in infection rates in the local area and other government data caches. But some users have complained about a lack of transparency in how the codes are determined. Users are neither given an explanation for the code they receive nor are they able to dispute it. They’re stuck at home indefinitely, waiting for their codes to turn green.
Emergency tools are here to stay
In a recent interview with Vice, Edward Snowden had this to say about the introduction of emergency measures to respond to COVID-19:
“Do you truly believe that when the first wave, this second wave, the 16th wave of the coronavirus is a long-forgotten memory, that these capabilities will not be kept? That these datasets will not be kept? No matter how it is being used, what’s being built is the architecture of oppression.”
As China’s lockdowns ease and life slowly returns to normal, what becomes of the Alipay Health Code? Checkpoints have largely disappeared from many cities. In Shanghai, where enforcement of the code was lax to begin with, most residents have stopped using it. However, local authorities are turning the systems on and off as they please. For example, in Guangzhou, public transport checks were renewed in late April following a sudden uptick in local infections.
Enforcement of the health code will understandably ebb and flow while the threat of a second wave of COVID-19 infection remains. But Snowden is probably right that these surveillance capabilities will persist long after the crisis ends. For example, police in Hangzhou have discovered that the Alipay Health Code has collateral benefits, after the absence of a green code reportedly forced a murderer to hand himself in after 24 years on the run. Authorities may well turn their minds to repurposing the health code for other purposes. Analysis of the Alipay Health Code reveals a piece of programming that already appears to be sending a user’s location and identification number directly to police. Another tool to add to the techno-authoritarian toolkit.
Lessons for other countries
In Australia, the government has been careful not to introduce any COVID-19 measure that smacks of a police state. Prime Minister Scott Morrison has said that Australia’s COVID-19 responses will reflect Australia’s values, adding this disparaging comparison – “I mean, in China they were welding people’s doors shut”.
While nowhere near as intrusive as the Alipay Health Code, Australia has joined countries like Singapore, the United Kingdom, and the United States in launching a voluntary app, COVIDSafe. The app captures Bluetooth signals between phones and an algorithm then determines any close contacts with an infected person. COVIDSafe has been downloaded by over 5 million people so far, but not without new privacy laws and public assurances that the data won’t be shared with law enforcement.
Genevieve Bell, Director of the Autonomy, Agency, and Assurance Institute at the ANU, argues there should be sunset clauses for the technologies we build for this pandemic, “in particular when it comes to the private, intimate, and community data we might collect”. But ensuring that there are sunset clauses, and that emergency powers do not become permanent and normalised, requires vigilance and a healthy public debate about the bounds of what is acceptable in a crisis. Citizens – Chinese, Australian, or otherwise – can’t rely on their governments to be vigilant for them. After all, a politician never wastes a good crisis.