The Criminal Code Amendment (Using Technology to Generate Child Abuse Material) Bill 2025 creates new offences in the Criminal Code Act 1995 to criminalise the use of online services and emerging AI technologies for generating, distributing or training systems that produce child sexual abuse material, with penalties of up to 15 years’ imprisonment and narrowly tailored public-interest defences.
The amendments respond to the growing misuse of generative AI and similar tools to fabricate exploitative imagery of minors.
This Bill amends the Criminal Code Act 1995 by inserting Subdivision 474DA into Division 474. It introduces:
The Act will commence on the day after Royal Assent and has no net financial impact.
The bill urgently addresses a novel and severe risk: sophisticated AI tools can fabricate or normalise child sexual abuse material without any real-world victim at the point of creation, yet these images perpetuate harm, fuel demand for exploitative content, and can retraumatise survivors. Criminalising the misuse of technology to generate such material is necessary to deter offenders and close an enforcement loophole.
By targeting the ‘‘sole or dominant purpose’’ of software and data pipelines used to produce child abuse imagery, the law is narrowly tailored to catch illicit operators without sweeping in benign applications. Public-interest defences ensure that legitimate research, compliance monitoring and law-enforcement activities remain lawful and unobstructed.
Australia’s international obligations under the UN Convention on the Rights of the Child require strong measures to protect children from all forms of sexual exploitation. This Bill implements those obligations in a proportionate and effective manner, promoting overall welfare and safety of minors.
While protecting children is paramount, the bill’s broad definition of ‘‘sole or dominant purpose’’ may inadvertently criminalise legitimate AI research, open-source development and security testing, chilling innovation in a critical sector. The boundary between malicious and benign AI tools is often unclear in practice.
Enforcement agencies and courts may struggle to apply the ‘‘sole or dominant purpose’’ test consistently, leading to legal uncertainty and the risk of prosecutorial overreach. Narrow defences for approved research do not fully capture the diversity of academic and industry work that contributes to safe AI development.
The Bill also limits privacy and expression rights by authorising surveillance of data-collection and software-distribution activities that may be used for entirely lawful ends. A more precise drafting or clearer guidance on intent and permissible purposes would better balance child protection with innovation and human rights.
2025-07-28
House of Representatives
Before House of Representatives
CHANEY, Kate, MP
Unspecified
Criminal Law Reform, Science / Technology, Discrimination / Human Rights