Australia tells itself a comforting story about child protection. We tell ourselves children are only removed from their families when there is no other choice. We tell ourselves the system protects vulnerable children from harm. We tell ourselves the Stolen Generations belong to the past.

But the numbers cut straight through that story.

Aboriginal and Torres Strait Islander children are roughly 6 per cent of Australia's child population but they account for more than 42 per cent of children in out-of-home care.

That figure has risen every year, without exception, since the Bringing Them Home report in 1997.

Australia apologised in 2008 for the forced removal of Aboriginal children. Parliament stood still. The country wept. Politicians promised a different future.

More Aboriginal children are being removed now than when that apology was delivered.

At some point, the country must stop pretending this is accidental.

The uncomfortable truth is Australia increasingly responds to Aboriginal poverty and social distress by taking children, rather than addressing the conditions destroying families in the first place.

Overcrowded housing becomes 'neglect'. Poverty becomes 'parental failure'. Intergenerational trauma becomes 'risk'. The state neglects to improve the conditions that would hold families together (housing, healthcare, economic security, infrastructure) but then turns up at the door with power to investigate when those same families struggle.

That is the real architecture of child protection in Australia. Not a rescue system. A damage management system, positioned at the end of a long chain of institutional failures it did not cause and is not designed to fix.

Consider what mandatory reporting - the legal obligation on teachers, health workers, police and others to notify authorities of suspected child abuse - actually measures. It measures visibility. A family experiencing overcrowding, poverty, family violence and untreated mental illness in a remote Aboriginal community serviced by government housing, government clinics and regular police patrols will generate far more system contact than a family experiencing identical stresses in a nice middle-class suburb.

The reporting apparatus is not distributed evenly across Australian society. It is concentrated, by design and history, in Aboriginal communities. What the statistics record is not simply harm. It is the intersection of harm with institutional surveillance. Australia has rarely examined that distinction honestly, because doing so would require asking what the surveillance is for.

After the death of five-year-old Kumanjayi Little Baby, the NT Government appointed former NSW Police Commissioner Karen Webb to lead an independent review, and has proposed amendments to the Care and Protection of Children Act that would place "child safety" considerations above cultural placement protections.

On the surface, this sounds self-evident. Of course, child safety comes first. But that framing and the speed with which it forecloses further inquiry reveal precisely how shallow Australia's understanding of Aboriginal child wellbeing remains.

For Aboriginal children, culture is not separate from safety. Family, kinship, language, belonging and connection to community are not optional extras to be added once physical safety has been secured. They are constituent elements of what safety is. The research on this is clear: strong cultural identity is among the most significant protective factors for Aboriginal young people, associated with lower rates of depression and anxiety, self-harm and suicidality, stronger educational engagement and greater resilience across multiple domains. This is not sentiment. It is epidemiology.

The Aboriginal and Islander Child Placement Principle, developed in response to the Stolen Generations and embedded in the child protection legislation of every Australian jurisdiction, exists because separating Aboriginal children from kin, culture, community and Country causes documented, measurable, multigenerational harm.

No serious person argues that children should remain in violent or genuinely dangerous homes. Some children absolutely need protection, and workers who provide it deserve support and adequate resources. But Australia continues to speak as though removal itself constitutes protection, as though the crisis ends at the front door of care. The evidence says otherwise. Too many Aboriginal children move from unsafe homes into unstable systems: multiple placements, separation from siblings, disconnection from culture, school exclusion, contact with the youth justice system, and eventually, in too many cases, adult incarceration. The state intervenes with considerable force at the point of removal and with considerably less force afterwards. And still the numbers rise.

Governments once used the language of "protection" and "civilisation" to justify removing Aboriginal children from their families. Protection legislation operating across every state and territory from the 1880s through to the 1960s held legal authority to take children without parental consent, without court oversight, and with the explicit purpose of eliminating Aboriginal cultural identity across generations.

The stated rationale was welfare. There was cultural destruction.

Today the language of "risk", "safety", "best interests" and "permanency" has been modernised. But the state's belief that it holds superior authority over Aboriginal families has not changed. Contemporary child protection workers are not the moral equivalents of Protection Board officers. Most are working under impossible conditions with inadequate resources. The problem is not the character of individuals inside the system. It is the logic of the system itself.

The deeper question is not why the system fails. It is why Australia keeps expecting a system weighted towards late crisis intervention to produce outcomes that can only be generated by early structural investment. Australia has consistently chosen the latter - it is cheaper at the moment of decision, more politically legible, and does not require the state to acknowledge its own role in the conditions being managed.

After every tragedy, governments promise reform. Thresholds tighten. Intervention expands. More Aboriginal children enter care. Another inquiry convenes. The cycle repeats with the dependable rhythm of a system that has mistaken its own activity for progress.

Removing children does not build houses. Removing children does not reduce poverty. Removing children does not heal trauma, close the gap between Aboriginal and non-Indigenous health outcomes, or address the structural underinvestment in remote communities that has persisted across governments of both political persuasions for generations.

And until Australia is willing to invest in Aboriginal families as seriously with the same resources, the same urgency, the same political commitment as it invests in removing Aboriginal children from them, this cycle will continue with the same outcomes it has always produced.

Dr James Beaufils is a Senior Research Fellow at the Jumbunna Institute of Indigenous Education and Research and the Faculty of Law University of Technology Sydney (UTS). He is a Gundungurra man from the Pejar area and Kanak from New Caledonia.

Dr Todd Fernando is a 2026 Paul Ramsay Fellow, senior researcher, and policy strategist focused on social justice, systems reform, and structural change in Australia.