Child abuse survivor launches SAFE child safety early-warning system

Ruby O’Rourke, a survivor of severe childhood abuse, has launched SAFE, a child-safety intelligence platform designed to detect behavioral warning signs earlier across Australian child-facing services. The system has already been used in more than 1,000 services and assessed records for more than 68,900 children, as Australia faces scrutiny over fragmented child-safety response systems.

Why it matters: - SAFE is designed to spot patterns of child distress before they become formal child-protection cases. - The platform targets Australia’s childcare trust crisis by giving educators a structured way to connect small behavioral signals over time. - Early detection could help child-facing services move from reactive notes and hindsight to earlier intervention.

What happened: - Ruby O’Rourke launched SAFE, a child safety intelligence infrastructure platform, in Sydney on August 5, 2026. - O’Rourke built the system after surviving forced adoption, sustained violence and near-fatal abuse as a child. - O’Rourke said the system is the answer to a childhood in which warning signs were missed.

The details: - SAFE lets educators record everyday observations that may look minor on their own but can indicate risk when combined. - The platform is built to flag signs such as withdrawal, appetite changes, unexplained fears and altered attendance patterns. - SAFE uses machine learning and natural language processing to analyze fragmented observations across time and locations. - The system surfaces children who may need attention before concerns would normally trigger a formal report. - SAFE was developed over 12 years in partnership with CSIRO/National ICT Australia, the University of Queensland and the Child Abuse Prevention Service. - The platform has been used across more than 1,000 Australian child-facing services. - SAFE includes assessment records for more than 68,900 children across every state and territory. - The platform has achieved about 97% accuracy across government pilots. - SAFE is machine-assisted but human-governed. - SAFE does not make child protection decisions. - Trained child protection professionals review and validate every behavioral signal before any action is considered. - O’Rourke said children often communicate distress through behavior before they can explain it in words. - Productivity Commission reports indicate Australia still relies heavily on fragmented notes, thresholds and hindsight when children are at risk.

Between the lines: - SAFE is positioned as a missing intelligence layer between frontline concern and formal child-safety response. - The launch reflects a broader critique of systems that expect educators to identify danger from isolated incidents instead of patterns. - O’Rourke’s personal history gives the project emotional force and a clear use case: the signals were there, but no adult connected them. - The company is also arguing that other high-risk sectors already use early-warning infrastructure, while child safety has lagged behind.

What’s next: - SAFE aims to extend beyond early childhood settings into schools, sport and broader human services. - The company is seeking interview requests through safeinc@thirdhemisphere.agency. - SAFE will likely continue positioning itself as a prevention tool for organizations trying to reduce escalation and improve early intervention.

The bottom line: - SAFE turns everyday observations into structured child-safety intelligence, with the goal of identifying risk earlier and preventing abuse from going unnoticed.