Login
iscover
Share this article on:
Powered by
MOMENTUM
MEDIA
Powered by MOMENTUMMEDIA
Feds and academics will develop a new tool to combat the production of AI-generated child abuse material, extremist propaganda, and malicious deepfakes.
The Australian Federal Police (AFP) and Victoria’s Monash University are teaming up to fight cyber criminals through what the AFP is calling a “dose of digital poison”.
The pair are working together via the AI for Law Enforcement and Community Safety (AiLECS) Lab to create a broadly capable disruption tool capable of slowing down and eventually stopping criminals from using AI-related technologies to create malicious deepfakes and child abuse material.
This data poisoning tool, known as Silverer, has already been in development for 12 months and is capable of causing the production of skewed or corrupted images.
According to AiLECS researcher and project lead, PhD candidate Elizabeth Perry, Silverer’s name refers to how mirrors are made.
“In this case, it’s like slipping silver behind the glass, so when someone tries to look through it, they just end up with a completely useless reflection,” Perry (pictured) said.
“Before a person uploads images on social media or the internet, they can modify them using Silverer. This will alter the pixels to trick AI models, and the resulting generations will be very low-quality, covered in blurry patterns, or completely unrecognisable.
“Offenders making deepfakes often try to use a victim’s data to fine-tune an AI of their own; Silverer modifies the image by adding a subtle pattern to the image which tricks the AI into learning to reproduce the pattern, rather than generate images of the victim.”
AFP Commander Rob Nelson noted that while such technology is in its infancy, the law enforcement possibilities “showed promising early results”.
“Where we see strong applications is in the misuse of AI technology for malicious purposes,” Commander Nelson said.
“For example, if a criminal attempts to generate AI-based imagery using the poisoned data, the output image will be distorted or completely different from the original. By poisoning the data, we are actually protecting it from being generated into malicious content.
“A number of data-poisoning algorithms already exist, and as we see in other cyber security areas, emerging methods to avoid them appear quickly soon after.
“We don’t anticipate any single method will be capable of stopping the malicious use or recreation of data; however, what we are doing is similar to placing speed bumps on an illegal drag racing strip. We are building hurdles to make it difficult for people to misuse these technologies.”
The AFP has noted a sharp increase in the amount of AI-generated child abuse material being created and has already made several arrests this year regarding the alleged possession or creation of such material.
“Currently, these AI-generated harmful images and videos are relatively easily created using open-source technology, and there’s a very low barrier to entry for people to use these algorithms,” digital forensics expert and AiLECS co-director Associate Professor Campbell Wilson said.
Commander Nelson said he hoped that the Silverer research project would eventually create a tool that “ordinary Australians” would find easy to use.
“We urge the public to consider poisoning images at risk of being manipulated by criminals for deceptive purposes,” Commander Nelson said.
“A dose of data poison will make it significantly harder for criminals to distort reality with artificial intelligence.”
David Hollingworth has been writing about technology for over 20 years, and has worked for a range of print and online titles in his career. He is enjoying getting to grips with cyber security, especially when it lets him talk about Lego.
Be the first to hear the latest developments in the cyber industry.
