Japan will soon join its more technologically aggressive peers in Asia with its own contact-tracing app aimed at mitigating the spread of COVID-19, as the nation inches its way back to a post-emergency life of normalcy.
Minimally intrusive and conditional on consent, the digital contact-tracing technology Japan is deploying is nowhere near as authoritarian in nature as those employed by countries such as China, South Korea, India and to a lesser extent, Singapore. But despite the privacy protections built into the platform, which the government is touting, the app’s effectiveness has been thrown into doubt, with some critics suggesting it is more of a security blanket than a powerful ally in tracing infection pathways — at least for now.
The app’s effectiveness is not the only source of controversy over the soon-to-debut app. As much as the built-in privacy protections are being extolled, the app, officials concede, is not entirely risk-free and could potentially clue in users on the identity of coronavirus patients they came into contact with.
Employing a technology provided by internet giants Apple Inc. and Google Inc., the state-developed app automatically exchanges Bluetooth signals between users. It will then at a later time send “exposure notifications” to those who came into close contact with patients who tested positive for the disease caused by the virus.
The way the system works is that the smartphone of one user will obtain and preserve the encrypted, randomly allocated codes of other users whenever they are within a meter of each other for more than 15 minutes. If one later tests positive for COVID-19 and agrees to confirm their infection status using the app, other users who have come into contact with them over the past 14 days will be notified.
By doing so, the app — rather than help the government identify patients and enforce quarantines — is designed first and foremost to notify users of their own risk and help them exercise caution so they can protect themselves and others.
With inclusiveness as one of its chief concepts, the app, which will be developed and run by the health ministry, will be available in multiple languages and released later this month.
The key to the app’s success lies in penetration. Recent analysis by a research team at the University of Oxford in England showed that roughly 60 percent of the population would have to use a contact-tracing app for it to be effective in controlling an outbreak.
To convince as many people as possible to download it, the Japanese government has gone out of its way to promote its noninvasive nature to dispel concerns that the tracing app will become another surveillance tool.
Unlike apps developed overseas, Japan’s doesn’t allow the government to centralize data, use GPS tracking or harvest phone numbers, names and other personal information — making it difficult to use as a surveillance tool. To top it off, the app’s use is contingent on the approval of users every step of the way.
Ironically, the consent requirement also means the app is “significantly at the mercy of users’ cooperation,” ultimately setting a high bar for widespread adoption, said Keio University law professor Tatsuhiko Yamamoto, who is on the panel tasked with scrutinizing the app.
For the app to function, “users must agree to first download it, and then to input their infection status should they test positive. This two-step procedure of consent is the biggest hurdle to people’s participation,” he said. In addition, they must activate Bluetooth and leave it on constantly for the app to work.
If history is any indication, achieving a penetration rate of 60 percent will be a herculean task, with Singapore often cited as a cautionary tale.
In Singapore, citizens were encouraged to install the noncompulsory app TraceTogether, which, like Japan’s envisioned program, exchanges Bluetooth signals with nearby mobile phones.
But concerns over data hacking and potential abuse of power by authorities — coupled with complaints of quickly draining batteries — have reportedly kept the download rate to an underwhelming 25 percent.
Moreover, with the city-state wrestling with a new wave of infections, authorities have changed tack, rolling out and mandating SafeEntry, a check-in system powered by QR codes, in many public spaces.
“That a state as small and authoritarian as Singapore, where the prime minister strongly exhorted the public to install the app, was able to achieve an adoption rate of just about 20 percent, and eventually forced to switch to a different method, speaks volumes about the difficulties that lie ahead for Japan,” said Hiroaki Miyata, a professor at Keio University’s Health Policy and Management School of Medicine.
It remains to be seen how willing the public will be in installing the app. Experts say a growing sense of complacency induced by the flattened infection curve in recent weeks suggests people are less likely to be motivated to use it than they would have been a few months ago, when the nation was in the grip of the virus.
But even if it takes hold, the app is unlikely to prove a replacement for traditional contact-tracing by public health centers.
“The best the app can do is to supplement” human-led contact-tracing, Miyata said. “Just because you download this app doesn’t mean that your safety will be guaranteed or every single route of infection will be identified,” he said, adding that until it gains traction, the app will likely remain something of an “amulet.”
Although much more innocuous than those in use overseas, the Japanese app does have its share of risks.
These include the possibility of a COVID-19 patient being identified by friends or acquaintances — a risk that might scare away prospective users in a nation where a litany of discrimination incidents against people merely associated with medical work, infected or not, have made headlines.
Exposure notifications will be sent in a way that not only keeps all parties anonymous but also withholds the specific times and places where the encounters occurred. Even so, a receiver who has interacted with only a few people over the past 14 days, for example, might be able to guess with relative ease who the infected contact is.
“That’s the kind of risk that even the least invasive app cannot eliminate,” said an official from a government IT task force debating the app. “If even this kind of risk has to be ruled out, then there is nothing an app can do,” he said.
Another downside is the fallibility of the exposure notification system.
Since the recognition of close contact is reliant solely on time and distance as measured by Bluetooth, smartphones with the app will start exchanging signals even if the device is left unattended.
This includes through partitions and screens, such as the clear plastic ones recently set up to block droplets expelled by customers and cashiers, and even thin walls.
“The important thing to remember is that the exposure notification is not completely accurate,” Yamamoto of Keio University said.
“The notification must be worded in a way that assures receivers that their close proximity to patients is not established yet, otherwise they might panic or overreact based on false assumptions,” he said.