“Deepfakes can be used to cause serious harm, especially when used to create sexually exploitive material of children and adults,” writes Senior Social Justice Advocate, Dr Mark Zirnsak.
By Dr Mark Zirnsak
“Christians believe human beings are created in the likeness and image of God, and they look to the example of Jesus to live their lives. All the recorded encounters between Jesus and children were kind, gentle and respectful. Children were central to the new social order Jesus initiated.”
World Council of Churches, ‘Helping Children Out of the Shadows and into the Light. Resources for Spiritual Life Addressing Sexual Violence Against Children’, May 2020.
The development of artificial intelligence (AI) is creating new words. One of those is “deepfakes”. As defined by the Australian eSafety Commissioner, a deepfake is:
“A digital photo, video, or audio file of a real person that has been edited to create an extremely realistic but false depiction of them doing or saying something they did not actually do or say. Deepfakes are created using artificial intelligence software that currently draws on a large number of photos or recordings of the person to model and create content.”
Deepfakes have been used for humour, creating parodies and memes, for entertainment.
However, deepfakes can be used to cause serious harm, especially when used to create sexually exploitive material of children and adults. In February, UNICEF expressed concern at the increasing use of AI to produce sexualised content involving children. In a UNICEF, ECPAT and INTERPOL study across 11 countries, at least 1.2 million children disclosed having their images manipulated into sexually explicit deepfakes in the past year. In some countries, this represented as much as 1 in 25 children.
In March, the UN expressed concern that deepfake abuse is part of a much broader pattern of digital violence targeting women and girls. They pointed out that sexual deepfake videos make up 98 per cent of all deepfake videos online, and 99 per cent of them depict girls or women.
UNICEF expressed concern that too many AI models are not being developed with adequate safeguards.
In June last year, the Australian eSafety Commissioner raised concerns that naked deepfakes of students and teachers were an increasing problem at schools. They were being used as a form of bullying or as part of deliberate image-based abuse, causing very real emotional and psychological harm. Victims of deepfake abuse can experience humiliation, shame, distress, fear, anger and confusion.
The Australian Parliament took an important step by criminalising the non‑consensual transmission of AI‑generated sexually explicit material via the Criminal Code Amendment (Deepfake Sexual Material) Act 2024 (Cth). However, regulation gaps remain in relation to sexually explicit deepfake material, and the regulation of non-sexual deepfake material is addressed only indirectly through older laws not designed for AI. Australia’s legal framework requires broader reforms to address the full spectrum of deepfake harms.
In January, concerns were raised that the Grok AI chatbot on Elon Musk’s social media platform X was being used to create and share non-consensual naked images of people and child sexual abuse material. It was reported that Grok was generating 6700 images of undressed children and adults an hour.
There is a need for technology corporations to have a duty of care to proactively detect and remove deepfake material. AI models should be developed to be safe by design, limiting their ability to generate harmful deepfake material.
The Synod’s Justice and International Mission Cluster has produced resources for anyone who wants to advocate for measures to reduce the harm caused by deepfakes. If you are interested in obtaining copies of the material, email jim@victas.uca.org.au
Dr Mark Zirnsak is the Synod’s Senior Social Justice Advocate
