All posts

Protech Progress Update

September 5, 2023

Several important steps have been achieved for the Protech research project. The EU-funded pilot study aims to develop innovative technology that will help stop the viewing of child sexual abuse material.

Since the project launched in March 2023, Protech consortium partners have been working towards the creation and deployment of Salus, a safety app designed to stop users from being able to see sexual abuse imagery on their electronic devices.

The two-year project will test the app to explore whether it could be part of an effective EU-wide prevention programme.

The distribution and trade in horrific child sexual abuse images and videos is rampant on the internet. In 2022, the Internet Watch Foundation – a research partner in the Protech study – removed more than 250,000 webpages containing criminal child sexual abuse content. Each webpage could contain thousands of images, amounting to millions in total.

Preventing people from being able to see this criminal content could help stop the repeated traumatisation of child sexual abuse survivors who suffer knowing that images of their abuse continue to be shared online.

Researchers have now conducted literature reviews to outline what is known about why people want to view child sexual abuse material and to better understand how artificial intelligence is currently being used to detect and remove criminal imagery from the internet.

Many participants will be involved in the study, including more than 50 professionals and at least 180 users over an 11-month period. As such, vital ethical approvals have been ensured to allow the research to be undertaken.

An advisory board has been established to help guide the research project and provide independent and expert advice.

To help build the app with the end users in mind, data collection is now under way across four countries – Germany, Netherlands, Belgium and the UK. This includes focus groups with practitioners who work with individuals at risk of viewing child sexual abuse material, as well as interviews with potential offenders themselves.

The information gathered from these interviews will be analysed and used to inform what the Salus app will look like and how it will function.

Designers behind Salus have been training the machine learning technology to accurately identify child sexual abuse material, using data sets of confirmed criminal content, including both still images and video.

The machine learning model has undergone rigorous testing and its performance measured against multiple test data sets using over a million inputs.

Front-end development will now consider feature specifications and how participants in the pilot study will use and navigate around the app.

Work is continuing apace on this project and important next steps include the release of the Salus prototype for user testing and the recruitment of participants for the pilot stage to test the prototype.