The Australian government has introduced its digital duty of care legislation to parliament, a bill that includes an “off switch” allowing social media users to opt out of algorithmic recommendation systems. If passed, Australians could choose to see feeds populated only by accounts they follow, a move welcomed by online safety advocates.
This approach echoes a model that one of the world’s largest platforms has run for over a decade: WeChat in China. Its long experience offers Australia and others a preview of what a feed unshackled from algorithms might look like — and it’s not all good news.
WeChat’s efficiency-first design
Launched in 2011, WeChat now has over one billion users globally, making it a primary social media service for Chinese speakers worldwide. The platform’s design logic was simple: WeChat was meant to be an efficiency tool. Unlike its American or Chinese counterparts, which use algorithmic recommendations to keep users engaged, WeChat is designed for users to open the app, do what they need, close it, and put their phone away.
Algorithmic curation played a minimal role until the introduction of the short-video function “Channels” in 2021, which competed with Douyin (the Chinese version of TikTok) and RedNote (Xiaohongshu) in the rising short-video market. Yet WeChat largely remains focused on efficiency, offering a “super app” suite of services — finance, commerce, publication, entertainment, and communications — within a single platform to make it integral to people’s lives.
The click-driven race
Our research project, RECapture, has studied WeChat Official Accounts targeting Chinese-Australians since 2019. Our data offer a longitudinal study of monitored content. We found that the absence of algorithmic curation gives publishers significant autonomy to generate revenue independently of the app. But within the platform, creators still fiercely compete for user attention, because users search for content they’re interested in and subscribe to accounts they find most compelling.
This competition has manifested as increased publication speed. Outlets, restricted by their own budgets, produce low-quality, clickbait content, often compiled from multiple fragmented and unverified sources, to attract subscribers and maximise user engagement. Increasingly, this race is accelerated by the use of generative artificial intelligence (AI).
The logic is simple: publish early and often; the clicks roll in throughout the day. Hit the right publication timing, the so-called “golden window”, and clicks spike further. Run a striking headline, and clicks climb yet again. This is the strategy to maximise a publisher’s chances of appearing in a user’s search query and capturing their attention. It’s clicks, clicks, clicks (or taps) that matter, as this is the metric to attract eyes to advertisements, and therefore money to the platform.
If the digital duty of care legislation passes in Australia and enough people choose to opt out of algorithms, we might see a similar increase in poor quality content here. It’s also possible the laws won’t deliver what they promise, because we don’t really know how digital platforms curate users’ subscribed content. We need the platforms to be more transparent about this.
My feed, my responsibility?
Under the promise of giving Australians more control over their online environment, the government effectively offloads its responsibility of care onto individual users. Exposure to harmful content implicitly becomes a consequence of the user’s failure to switch off their algorithm. This could let tech companies off the hook.
The legislation also reads as a compromised version of the abandoned Combatting Misinformation and Disinformation Bill. Rather than tasking law enforcement bodies with addressing the issue (which risked being entangled in freedom-of-speech debates), the government instead leaves citizens to manage the problems themselves. The government knows the product harms, so it is making sure to protect children through the under-16 social media ban, even while individualising choices for harmful digital products for everybody else.
Getting to the root cause
The Australian government has started viewing digital platforms more critically, a welcome change from laissez-faire attitudes of the past. However, the government tends to understand digital platforms as a bunch of functions and tools rather than products of technology business entities, with which the government nonetheless continues to do deals. This means the government operates on the assumption that switching off the “bad” ones, cutting off access, or introducing surveillance systems is a straightforward solution to the problems we face in our digital environment.
What this view overlooks is that harm stems from the business model of tech companies. As long as that business model remains unchanged, harms will not disappear. Being able to opt out of the algorithm will not shift the business model of content creators for the Australian market. And how the platforms themselves navigate this remains to be seen.
While the government’s legislation focuses on the problems of social media over the past decade or so, the underlying issues are deeper. The Australian experience with WeChat’s model suggests that simply switching off algorithms is not a panacea. It may reduce engagement-driven manipulation, but it could also foster a race to the bottom in content quality, as seen on WeChat. For a more effective approach, policymakers should consider addressing the incentives that drive platforms and creators alike, rather than just tweaking the interface.
