Australia's Digital Duty of Care: A Death Knell for Self-Regulation in Dating Apps

Match Group and Bumble launched Australia's voluntary dating safety code in October 2024 with considerable fanfare. Five months later, the Australian government has made its view unambiguous: self-regulation doesn't work. The proposed Digital Duty of Care, released for consultation earlier this month, would impose legally binding obligations on dating platforms to identify and mitigate foreseeable harms—a fundamental shift from asking nicely to demanding accountability.

The speed of this pivot is remarkable. Voluntary codes typically get at least a year or two to demonstrate effectiveness before regulators lose patience. Australia's decision to move to mandatory safety duties after just five months suggests either the government never genuinely believed industry commitments would deliver, or early evidence pointed to such negligible change that waiting became untenable.

The stakes for platforms

This isn't just another regulatory headache for compliance teams—it's a referendum on whether dating platforms can be trusted to police themselves on safety at all. Australia's rapid abandonment of the voluntary approach will embolden regulators in other markets who've been watching this experiment closely. The UK, EU, and possibly even US states now have a template for moving beyond voluntary frameworks.

Create a free account

Unlock unlimited access and get the weekly briefing delivered to your inbox.

Register free

No spam. No password. We'll send a one-time link to confirm your email.

For heterosexual women, harmful behaviour on dating apps isn't an edge case. It's the baseline experience.

More fundamentally, the 99% harm figure—even accounting for definitional breadth—exposes a truth the industry has spent years trying to soften. For millions of users, these platforms don't just occasionally fail at safety. They systematically expose women to psychological, emotional, and fear-based harms as a routine feature of participation.

What the research actually shows

The numbers from the Behavioural Insights Team study are staggering, though they require careful interpretation. According to the research, which surveyed 1,008 heterosexual women in Australia, 99% reported experiencing at least one harmful behaviour on a dating app in the preceding three months. That figure encompasses a spectrum—95% reported psychological or emotional harm, 89% experienced monitoring or surveillance behaviours, and 58% faced fear-based controlling behaviour including threats of physical harm.

Context matters here. The study's definition of "harmful behaviour" likely ranges from unwanted sexual messages to credible threats of violence. That breadth explains the near-universal incidence rate. But even at the severe end, the findings are stark: nearly six in ten women reporting fear-based control or threats isn't a marginal safety problem.

Perhaps more revealing is the reporting rate: just 22% of women who experienced these behaviours reported them to the platforms. That figure suggests either profound lack of confidence in platform response mechanisms, or—more troubling—such widespread normalisation of harmful behaviour that most women don't consider it worth reporting. Both possibilities indict the current approach to trust and safety.

Trust and safety teams at the major platforms will argue that defining "harm" so broadly makes the 99% figure misleading, and they'd have a point about precision. But the political reality is simpler: when a government-commissioned study shows near-universal harm among women, the case for mandatory intervention becomes politically irresistible regardless of definitional quibbles.

How the duty of care would work

The proposed Digital Duty of Care amendment to Australia's Online Safety Act shifts the burden of safety from users to platforms. Rather than relying on post-incident reporting and moderation, services would be required to proactively identify foreseeable risks and implement measures to reduce them before harm occurs.

The exposure draft was released on 8 September, with consultation closing today (22 September). Legislation is expected to reach parliament in October, moving at remarkable speed for regulatory reform of this scope. Dating apps fall explicitly within scope, alongside social media and other user-generated content platforms.

Critically, the duty applies to "foreseeable" harms—a legal standard that platforms will struggle to contest given the BIT research and their own incident data. Claiming you couldn't foresee that women face harassment, threats, or controlling behaviour on dating apps becomes untenable when government studies document 99% incidence rates.

Post-hoc content moderation won't satisfy a proactive duty of care. Platforms will need to demonstrate they've designed against known harms from the outset.

For operators, this means rethinking architecture. Platforms will need to demonstrate friction in messaging, limits on information disclosure, detection systems for coercive behaviour patterns, and meaningful consequences for repeat offenders. The voluntary code included many of these commitments on paper. The difference now is legal liability if they don't work.

The global implications

Australia has form as a regulatory testing ground for tech accountability. The country legislated to force Meta and Google to pay news publishers, introduced some of the world's strictest content moderation penalties, and is currently debating social media age verification requirements. Dating apps are simply next in the queue.

Other jurisdictions are watching. The UK's Online Safety Act already imposes proactive safety duties on user-to-user services, though dating apps received specific carve-outs on certain provisions. Australia's explicit inclusion of dating platforms—backed by research documenting endemic harm—will fuel calls to tighten those requirements.

The EU's Digital Services Act similarly requires risk assessment and mitigation, though enforcement remains nascent. What's different about Australia's approach is the directness. There's no regulatory forbearance period, no grace for good-faith efforts, no assumption that platforms will eventually get this right if given enough runway.

The harder question is whether mandatory duties can actually solve the underlying problem. Dating apps facilitate connections between strangers, often with romantic or sexual intent. Some level of unwanted attention, miscommunication, or boundary-testing behaviour may be inherent to that dynamic. Engineering out gendered abuse without fundamentally changing how the platforms work—heavy friction, verified identities, restricted communication—remains the unsolved challenge.

Meanwhile, South Australia is proposing to ban perpetrators of domestic and sexual violence from dating platforms entirely, representing another attempt to shift responsibility for safety away from users and onto those causing harm.

For Match Group, Bumble, and other signatories to the voluntary code, the message is unambiguous: demonstrating compliance through features and policies no longer suffices. Governments want proof that women's lived experience on these platforms is materially improving. Absent that proof, expect more markets to follow Australia's lead.

D