The news that Chinese-owned video-sharing app TikTok is not approved for use on devices owned by Australia’s Department of Defence, as the ABC reported today, is hardly a surprise.

Defence’s default position on what apps it allows on its work phones is, in effect, that all apps are banned unless there’s a need for them and they pass a security test. Unsurprisingly, seeing how our defence personnel put their own spin on the ‘Haribo Challenge’ is not a top priority for the department.

The news follows a warning in mid-December from the Pentagon to various US military branches to remove the popular app from their phones. The Australian decision, as with the ban on the use of Chinese-owned app WeChat almost two years ago, is a no-brainer.

Like most social media apps—or any other app, for that matter—TikTok and WeChat require their users to agree to a startlingly invasive list of permissions. A recent analysis of TikTok’s privacy and data-collection policies revealed that it demands a worryingly high level of access to systems and information on both Android and Apple devices.

These include requiring full access to the device’s camera and microphone (hardly surprising for a video-sharing app) and its contact list, as well as detailed location data using GPS. The GPS tracking is ‘surprising’, say the researchers at security company Proofpoint who analysed the app, ‘especially as TikTok videos don’t obviously display location information’.

The metadata harvested from these apps can paint a vivid picture of what military personnel are up to, as the 2018 Strava app debacle uncovered by my ASPI colleague Nathan Ruser showed. The Strava case demonstrated, too, that an app doesn’t need to be owned by a foreign adversary for it to inadvertently pose a serious risk to military operational security.

But when the app is owned by a company operating in an authoritarian country, as TikTok’s owner ByteDance is, there are added risks to users, whether they’re military personnel or not. It’s those risks that prompted a full-scale pushback from US authorities as TikTok’s popularity rose in America.

The committee on foreign investment there has said the app could pose national security risks for Americans and possibly be used to influence or monitor them. In November, the New York Times reported that the US government ‘had evidence of the app sending data to China’. In a recent lawsuit in California, the plaintiff has alleged that TikTok transferred vast quantities of her private and personally identifiable data to servers in China.

TikTok has denied that its data is being sent back to Beijing, stating that all US user data is stored in the US, with backup redundancy in Singapore. But, as Proofpoint’s researchers pointed out, there’s no information on TikTok’s website about where data on users from other countries, including Australia, is stored.

Even if Australian users’ data is being stored in Singapore, ByteDance’s engineers, who are based in Beijing, would still need to access that data in order to continue to improve the app, as David Carroll, an associate professor of media design at Parsons School of Design, has argued.

Once it’s there in Beijing, it could be easily accessed by the authorities. After all, as is now well known (thanks in part to the Australian government’s decision to ban Huawei from taking part in the rollout of the national 5G mobile network), China’s National Intelligence Law from 2017 requires organisations and citizens to ‘support, assist and cooperate with the state intelligence work’.

The pushback against TikTok, which this Australian Defence ban has now played a small role in, has reportedly prompted internal discussions at ByteDance as to their best course of action. The options being canvassed at the moment, according to Bloomberg, range from an ‘aggressive legal defense and operational separation for TikTok’ to ‘sale of a majority stake’.

ByteDance, it seems, is determined to put as much daylight as possible between the company and Beijing. TikTok CEO Alex Zhu says he would turn down a request from President Xi Jinping himself to censor content on the app—an entirely meaningless claim, given that’s not even remotely how China’s censorship system works in practice.

A recent TikTok transparency report tested out another bit of sophistry. The report claimed there were zero takedown requests from China, but five from Australia. However, given that the version of TikTok that Australia has doesn’t work inside the Great Firewall of China, that’s hardly surprising at all.

Defence has clearly not been fooled by such obfuscations. That the company is now entirely captured by Beijing’s censorship and surveillance apparatus is beyond doubt. Only months ago, for example, Uyghurs in Xinjiang were using Douyin, the version of TikTok used within the Great Firewall, to shine a light on the brutal surveillance state. Now, chillingly, those videos of despair have been expunged and replaced with shiny happy people holding hands.

Originally published 28 May 2019.

I recently sailed onboard the landing helicopter dock and Royal Australian Navy flagship HMAS Canberra from Vietnam to Singapore, as one of several academic sea-riders invited to observe Indo-Pacific Endeavour 2019 (IPE 19), which has now concluded after three months at sea.

A key question about this operation was what did the People’s Liberation Army Navy make of so much Australian activity in waters that China claims? Sailors tend to have a healthy professional regard for each other. The good news is that interactions between the RAN and PLAN remain cordial.

But we were followed at a discreet distance by a Chinese warship for most of the transit, both on the way up and back, despite the fact that our route didn’t take us near any feature occupied by Chinese forces, or any obviously sensitive areas. Bridge-to-bridge radio communications, I was told, while courteous, included requests for Australian warships to notify the Chinese in advance of any corrections to their course—something the RAN was not about to concede while exercising its high-seas freedoms.

The presence of our ‘trailing escort’—HMAS Parramatta apparently received similar attention from the PLAN—had no obvious effect on HMAS Canberra’s activities, including maintaining a brisk pace of flight operations. The army’s Tiger attack helicopters practised night flying and deck landings, as they start to develop a maritime role.

Some helicopter pilots had lasers pointed at them from passing fishing vessels, temporarily grounding them for precautionary medical reasons. Was this startled fishermen reacting to the unexpected? Or was it the sort of coordinated harassment more suggestive of China’s maritime militia? It’s hard to say for sure, but similar incidents have occurred in the western Pacific.

Since its first outing in 2017, IPE has become established as Australia’s roving vehicle for regional defence engagement, though it more closely resembles a set of separate port-based engagements than a rolling exercise at sea. The aim is to shape Australia’s strategic environment, with a heavy emphasis on defence diplomacy, to strengthen relations. Although delivered by a naval task group, the joint aspect of IPE 19 was emphasised by the fact that it was led by an air commodore.

Although IPE 19 was primarily focused on South Asia, for me this was an opportunity to experience a naval transit of the South China Sea (SCS), in the hope of demystifying that most commented-upon body of water. I’m grateful to the RAN for making that possible. While the transparency and openness I experienced onboard was impressive, the SCS remains something of an enigma.

The fact that IPE 19 incorporated a spur into the SCS is itself significant. Australia was literally going out of its way to include Vietnam in the activity.

The ADF’s military-to-military engagement with Vietnam was less evolved than with more familiar partners along the IPE route—but more interesting for being unfamiliar. One of Vietnam’s six Kilo-class submarines based at Cam Ranh Bay was conspicuously on hand to witness our departure. We then conducted a passage exercise with two Gepard-class frigates, the Vietnam People’s Navy’s largest warships. The Vietnamese went out of their way to engage in broader respects, including a rock-star welcome to ADF personnel on a school visit (many onboard ranked this as their personal highlight of IPE 19).

In addition, a small contingent of US Marines based in Darwin had embarked in Thailand. My first significant moment of IPE 19 came in Nha Trang (a popular recreation spot for US servicemen during the Vietnam War), where I witnessed probably the first US Marines, out of uniform but on active duty, to walk along that beachfront in nearly 50 years. Historical value aside, this was surely a significant defence diplomacy ‘effect’, sharpened by the fact that it was delivered, benignly, from an Australian amphibious assault ship.

It also says something that the US Marines were welcomed in Vietnam but didn’t go ashore in either Malaysia or Indonesia. Sensitivities—not only on the hosts’ part—can hopefully be salved so that the marines can in future interact with more regional partners, under an IPE aegis. It will take time for all concerned to become accustomed to US forces ‘plugging in’ to an Australian operational and diplomatic lead. But this kind of role reversal is in the long-term interests of the alliance, as Canberra steps up strategic engagement in the region.

My short time embedded with the RAN also made me reflect on how an inordinate academic and media focus on freedom-of-navigation operations, something I’ve contributed to, has distracted from more fundamental questions about Australia’s naval presence and willingness to operate in the SCS.

For a modest-sized navy, the RAN has been remarkably active across the SCS. HMAS Canberra transited from Cam Ranh to Singapore, accompanied by HMAS Newcastle (its final voyage before decommissioning). Meanwhile, HMAS Success (also about to be paid off) was participating in multinational ASEAN Defence Ministers Meeting (ADMM)-Plus exercises to the south. On our way to Singapore, we crossed paths with the Anzac-class frigate HMAS Parramatta, which was headed on a northeasterly track. HMAS Melbourne had passed through the SCS and Taiwan Strait only a few weeks earlier, on its way up to Qingdao and points east.

The ubiquity of PLAN vessels shadowing other warships in the SCS suggests that China’s surface force has grown big enough to be able to ‘close mark’ at will. Our transit coincided with multinational naval activity elsewhere in the South China Sea, including a US FONOP close to Scarborough Shoal. The PLAN was simultaneously participating in ADMM-Plus exercises to the south, alongside HMAS Success.

The fact that Chinese warships are shadowing their marks for longer distances suggests that the PLA’s over-the-horizon surveillance capability is also maturing. The large facility at Fiery Cross Reef sits at a prime gatekeeper location in the Spratly Islands, from which PLAN ships on ‘picket’ stations can be sent to intercept other vessels.

Although IPE embodies non-warlike missions, including humanitarian assistance and disaster relief, our transit through the most hotly contested body of water in the Indo-Pacific brought home the contrast between a naval task group operating under peacetime conditions, scattering and then reforming, and the more challenging scenario of sailing a naval task force into harm’s way, in a decreasingly permissive environment like the SCS.

Australia’s navy has gamely demonstrated its willingness to remain present in the SCS, operating in several locations at once. China has equally underlined that it can detect foreign warships and has the capacity to deploy screening forces across the SCS. That’s a sobering development.

If Australia continues to operate in the SCS in the long term, as I hope it will, its ability to do so will depend increasingly on the support of its allies and partners. This is not a job to do alone. IPE represents a head start on developing the defence and security partnerships that Australia needs to invest in for a less benign future.

Countries’ 5G networks have become the subject of intense debate—and the Australian government can take some of the credit for that. Its decision to exclude high-risk vendors in August 2018 was marked by a refreshing openness about the grounds on which it was made. The thinking about risk and security was very similar to that in the EU security assessment of 5G released in October 2019.

Two core factors in these assessments are that 5G providers that have legal and other imperatives to cooperate covertly and deeply with their home governments pose a greater security risk than providers without these obligations and imperatives—and that governments do not want to be beholden to the governments of states with which they do not share values or interests. Both these factors apply in the case of the Chinese state and its two 5G ‘national champions’, Huawei and ZTE.

While there has been a burgeoning and critical debate inside and between governments, there’s been strikingly less talk about what the future 5G world means for ‘big tech’ in China and the US—the two homes of the globe’s tech giants.

That’s surprising, because arguably the strategic and economic competition between Xi Jinping’s China and Donald Trump’s America has at its centre a contest over advanced technology and the future of global data and communications. Xi’s China Dream can’t be achieved without China’s big tech firms dominating their US counterparts globally, because gaining strategic and economic power depends so heavily on having technological ascendancy, especially in data and communications.

But even without the state-on-state great-power competition, Alphabet (Google’s parent company), Amazon, Apple and Microsoft all have existential stakes in the future of global communications. The odd thing is they don’t seem to be acting like it.

If, as some experts predict, Huawei and ZTE win the race to build and operate much of the world’s 5G infrastructure, US tech firms will be dependent on them to transmit their services and products. American big tech will live in a Chinese walled garden of technology and be subject to the curators’ choices about what it can and can’t do, at least in major markets like Southeast Asia, Africa and chunks of Europe.

On the flip side, China’s non-5G tech giants Baidu, Alibaba and Tencent (BAT) will ride on the digital infrastructure of Huawei and ZTE. Given the Chinese Communist Party’s digital authoritarian model of fused state–corporate cooperation, this puts BAT in an advantageous position—not just in the China home market, but wherever the Chinese 5G champions provide the digital infrastructure.

The worst-case scenario for big US tech is that Apple, Facebook and Google in effect become apps inside a Chinese global information ecosystem. Chinese state policies and BAT, Huawei and ZTE would decide which of their products and services can be hosted on Chinese digital infrastructure and services. That looks like a very bad business plan, unless entering bankruptcy or becoming CCP-compliant in your global operations is the end goal.

So far, Apple’s and Google’s forays into 5G have been about 5G smartphones—which will also ride on the underlying 5G infrastructure.

So why are these US tech giants seemingly snoozing at the wheel on 5G and future communications infrastructure? Maybe they’re like great white sharks, lurking deep but ready to break the surface and show us that they knew what they had to do all along. But there’s no sign of that.

Maybe they’re planning to get a jump on 6G, thinking that new communications concepts and systems—like cheap and ubiquitous constellations of small satellites—will deliver a new source of communications business advantage that undercuts terrestrial 5G networks.

Maybe the real reason Google insisted on continuing to develop the now apparently defunct model of its search engine—Dragonfly, which was being designed to comply with Chinese state censorship—is because it’s preparing for the corporate dystopia of the future, not just trying to find a way to get bigger in the Chinese market.

Maybe they all think the competition between the US and China is just the business of governments and therefore doesn’t concern them. If so, they’re failing to see that the same factors which are driving renewed great-power competition mean that they too are in strategic and economic competition not just with BAT, Huawei and ZTE, but with their chief supporter, the CCP.

Or maybe they’re stuck in a mindset that governments around the world already know is dead: the idea that globalisation is a process that ‘de-states’ the world, rendering state boundaries and powers less relevant. According to this long-gone vision of globalisation, companies and consumers could be ‘country agnostic’ about where services and products came from because global supply chains and capital would move and reconfigure in ways that had no little or connection with individual nations.

That was the arc of logic operating in much of the world politically and economically until the return of assertive authoritarian power—Russia, and now, most particularly, the Chinese state under Xi. And the truly globalised world was the dominant paradigm when many of the leaders of US big tech rose to prominence, so it wouldn’t be surprising if they’re finding it hard to let go. For all their genius, they too are products of their environment and their times.

Xi and great-power competition—along with ‘Made in China 2025’, the Digital Silk Road and military–civil fusion—have killed that vision.

There’s a way out of the Chinese walled garden for Alphabet, Amazon, Apple and Microsoft. These companies have enormous capital and human capital strengths. And they have potential 5G, data and communications partners in places other than China—Samsung in South Korea; Japanese tech giants like Sony, NEC and Fujitsu; Europe’s Nokia, Ericsson and even Siemens; and US partner-competitors like Qualcomm and Oracle (as well as India’s growing tech world).

Certainly, democratic governments from Canberra to Washington, Seoul, Berlin, Delhi and Tokyo all have deep interests in creating policy and regulation that supports alternatives to a unitary Chinese tech future. Fraying international cooperation, the disdain for Trump in much of Europe, and Trump’s own disdain for alliances and partnerships make these interests harder to give effect to. Yet letting those difficulties prevent positive action to shape our future communications world would be a defining and generational error for each of the democratic world’s leaders.

Big tech companies and their home governments outside the authoritarian world need to reset their relationship from a contested one around privacy, data and regulation to one that supports government and corporate strategic and competitive interests. That seems a remote prospect in the world we’re in right now, but what seems remote has a habit of happening when the need is sufficient.

Even without supportive policies and regulation from their home governments, it makes sense as a simple matter of future corporate existence, success and competition for the US tech giants to be able to live on alternative digital infrastructure to that brought to them by China Tech Inc.

Who from Alphabet, Amazon, Apple and Microsoft will be the first to get this party started?

Make no mistake about it; Hongkongers’ overwhelming vote on 24 November was a clear and unequivocal repudiation of Beijing and its appointed government here in the territory. It was a rejection of police abuses and repression, and an irrefutable popular endorsement of the demands of the protesters, if not their violent tactics.

Beijing can try to dismiss or downplay the results. The Chinese Communist Party’s media mouthpieces can hyperventilate about ‘foreign interference’, ‘intimidation’ by protesters and mythical voting irregularities. And the humiliated, routed pro-China parties here, seeing their ranks on the local councils decimated, can try to spit mathematical hairs, arguing, for example, that the pro-democracy forces ‘only’ won 60% of the vote—as if 1.6 million people voting and a 71.2% turnout amount to anything other than an electoral landslide.

But all spin aside, there is no mistaking that the pro-China bloc in Hong Kong took a drubbing of monumental proportions. The only question remaining is how Beijing’s communist rulers and their anointed minions running Hong Kong will respond to this expressed will of the true silent majority. The choice is either reform or further repression.

For many who know China and its rulers, the clear answer is repression—and they recall the brutal crackdown of June 1989, when troops of the People’s Liberation Army crushed the last major pro-democracy uprising by massacring hundreds, if not thousands, of young protesters who had occupied Tiananmen Square. The Tiananmen massacre, which Beijing has attempted to erase from the collective consciousness ever since, stands as a testament to how far a tyrannical autocratic regime will go to maintain power.

But another, more recent event in China offers a parallel path, showing that when faced with a challenge to its authority, China’s rulers can also be adept at accommodation and compromise to defuse a crisis. Instead of Tiananmen Square in 1989, I am thinking of Wukan in 2011.

Wukan is a small fishing hamlet in southern Guangdong Province where, in late 2011, villagers fed up over the seizure of their land revolted, chased communist party officials and police out of the village, and took control of the local government.

Rather than move in with force to crush the village uprising, the central government eventually sent in the provincial party secretary of Guangdong, Wang Yang, who was seen at the time as a leading economic reformer. Wang made significant concessions to resolve the weeks-long uprising, including freezing some of the land deals at the heart of the dispute, releasing jailed villagers from custody and sacking some recalcitrant local officials.

In other words, the central authorities admitted that the local officials had erred, launched an investigation and basically ceded to the protesters’ demands.

It was a pattern I had seen often while covering China as a correspondent from late 2009 until 2013. For example, when faced with a series of labour strikes and widespread worker unrest in 2010 and 2011, Chinese authorities responded first with the police, sparking violence clashes, but eventually urged localities and provinces to raise the minimum wage for workers. The unrest was soon defused.

In 2009, a 19-year-old man named Sun Zhongjie, working as a driver for a construction company, was fined for operating an ‘illegal taxi’ after giving a lift to a man who flagged him down on the road. The Shanghai traffic police were running a sting operation against unsuspecting motorists, and Sun—fined the equivalent of $2,000 and fired from his job—made a dramatic public declaration of his innocence by chopping off his little finger. After Sun’s case sparked an uproar on social media, he won his case and didn’t have to pay the fine, and hundreds of other drivers who had been ensnared in the scheme received refunds.

Since the Tiananmen massacre and the advent of the internet, China’s authoritarian rulers have held to power with far more than just the barrel of the gun. They have become adept at keeping track of public opinion through a vast network of monitoring centres at universities and state-run news agencies, following all the trending discussion online and trying to steer local officials to defuse potential crises before they erupt.

Of course, there are major differences between the situation in China now and even a few years ago. President Xi Jinping, who has eliminated term limits, effectively allowing him to rule for life, seems far less interested in compromise and concessions than his predecessor Hu Jintao. As seen by his actions in Xinjiang with the internment of a million Uyghurs in concentration camps, Xi seems more interested in ruling through power and fear than negotiation.

Also, since that relatively open 2010–2013 period, Beijing’s stifling of the internet has been nearly complete. Weibo, once a freewheeling platform for debate, is a shell of its former self. Online dissent can be more quickly snuffed out. The media is more strictly controlled than ever. Even virtual private networks used to bypass the censors are more difficult to use.

But that crackdown on the internet and the media brings its own problems for China’s leadership— like leaving them unaware of the depth of popular discontent in Hong Kong.

There has been some informed speculation that China’s leaders, and particularly Xi, were unpleasantly surprised by the outcome of the Hong Kong vote and the overwhelming defeat of the pro-China camp. According to this widely held view, officials in Hong Kong were assuring Beijing that a ‘silent majority’ was opposed to the protests and backed the government and police.

One thing that has become certain over the years is that China’s rulers do not like to be surprised.

With the depth of dissatisfaction in Hong Kong now made abundantly clear at the ballot box, China and Xi now face a choice—whether to double down on repression or listen to the people and opt for a new course that will require some compromise.

The contours of the compromise have been apparent for months. Hapless Chief Executive Carrie Lam and her incompetent cabinet and advisers must go. An independent commission must be empowered to investigate the violence and excessive force used by the Hong Kong police against the protesters. A ‘truth commission’ must probe the causes of the unrest and offer amnesty for any protesters who didn’t injure anyone or cause serious property damage—and that would mean releasing most of the thousands arrested on ‘rioting’ charges.

And finally, Beijing needs to respond to the people’s clear aspirations for autonomy by announcing a relaunch of the long-stalled political reform process, with a promise to eventually allow all Hongkongers a vote on their leader.

Whether Beijing will choose the reform path remains to be seen. Like with the change to the opening and reform policy in 1979, China’s communist leaders have shown they can deftly shift positions and adapt when they need to. And in Hong Kong right now, shifting and adapting is what they clearly need to do.

China’s global disinformation campaign—which recently painted Hong Kong’s democracy advocates as violent and unpopular radicals—should cause concern for democracies around the world. In addition to highlighting the many ways in which technology can be used to suppress freedom of speech, China’s ascent to the global disinformation order demonstrates a deeper insidious trend: governments increasingly see social media as a powerful tool to manipulate public opinion both domestically and abroad.

At Oxford University, we have tracked how governments use social media to spread computational propaganda. Over the past three years, we have found evidence of disinformation campaigns run by state actors in more than 70 countries around the world. Many of these countries are authoritarian regimes that use networks of fake accounts to spread pro-government propaganda, drown out opposing voices, and threaten activists and journalists with hate and violence. But the case of China is particularly worrying.

The Chinese government has a long history of censorship and information control. Even before ‘fake news’ was weaponised by US President Donald Trump, Chinese officials used the term to crack down on political dissent and discredit opinions that challenged the position of government. And it has become well known that fake online commentators—employed by the Chinese government—are responsible for fabricating hundreds of millions of online posts on Chinese native platforms every year to divert criticism away from the state. But until recently, China’s disinformation efforts have largely remained within the boundaries of its Great Firewall.

Before the Hong Kong protests, Chinese foreign influence operations were small in scale and relatively unsophisticated. We know they used crudely automated accounts to attack political figures in Taiwan, and have targeted social media disinformation campaigns against Falun Gong and Tibetans in exile. But now—with Facebook, Twitter and Google taking down hundreds of accounts, pages and channels operated by the Chinese government—China has demonstrated its capability and willingness to target global audiences with disinformation.

China’s quick, calculated and deliberate rise to the global disinformation order has two implications for democratic societies.

First, despite the numerous countries experimenting with computational propaganda, China is poised to become the next disinformation superpower. China has quickly become a global leader in artificial intelligence. Combined with the vast amounts of data collected from China’s social credit system and critical infrastructure rollouts around the world, AI will boost the state’s ability to target, tailor and amplify disinformation campaigns.

Democracies already have to contend with disinformation from domestic sources, as well as a growing number of sophisticated actors using social media for foreign influence operations. China—along with its technological prowess—is a significant player in the disinformation realm.

Second, China’s disinformation campaign against protesters in Hong Kong shows that we are still stuck. We remain beholden to social media firms to tell us when there is trouble, and not all governments are committed to combating disinformation.

Major technology firms don’t always stand up to China, so it is a great sign that they have exposed such interference. But independent researchers are still on the outside, unable to look under the bonnet because the most valuable data about public life remains in private hands. The question of public funding for better elections administration is on hold in the US. Highly trusted public broadcasters in many countries, which effectively inoculate the public against a lot of misinformation, have shrinking budgets.

There are a host of good ideas for using social media to strengthen democracy. Facilitating ‘data donation’ by users, contributing data to public archives, and setting up industry-funded but independent auditing bodies would all help put civic life back into social media.

But until the technology giants become better corporate citizens and politicians commit to securing elections, voters in democracies will remain soft targets for computational propaganda—from a growing number of countries with increasingly sophisticated technologies.