Description

It’s been billed “the Netflix of AI.” A little-known Australian startup running an online service, boasting more than 20 million users worldwide. It promises to bring your fantasies to life with realistic AI companions you design and control. What reporter Josh Robertson discovered was something more disturbing: characters that appear as young as 14 and 15 years old, presented for adult role-play, and a company unable to stop them appearing.

Summary

This investigative report examines an Australian AI startup's failure to prevent the creation and consumption of sexualized child imagery within its AI companion platform. The findings highlight significant gaps in safety guardrails for generative AI models and the challenges inherent in moderating decentralized, user-generated AI content. While the focus is on societal harms and ethics rather than existential risks, the investigation underscores the urgent need for robust Australian AI governance to address immediate safety failures in frontier-style technologies. The case serves as a critical example of how commercial AI deployment can outpace regulatory capture and safety enforcement within the domestic landscape.

Body