A new report by the non-governmental organization Human Rights Watch says that over 170 images and personal information of children from Brazil have been collected, without their knowledge or consent, and used to train AI tools.
The images have been collected from content published as recently as 2023, but there are also much older ones, from the mid-1990s. At the time, no one would have imagined that its content could be used to train artificial intelligence. Human Rights Watch claims that the personal information of these children, along with links to their photos, was included in LAION-5B, a dataset widely used by AI startups to train their tools.
See also: EU: Microsoft accused of collecting children's data
“Their privacy is violated when their photo is collected and scanned into these datasets. And then these AI tools are trained on this data and therefore can create realistic images of children,” says Hye Jung Han, children’s rights researcher at Human Rights Watch. “The technology has developed in such a way that any child who has any photo or video on the Internet is now at risk because any malicious user could take that photo and then use these tools to manipulate it however they want.”
LAION-5B has been used to train many artificial intelligence models, including Stability AI's Stable Diffusion image generation tool . The LAION-5B dataset is openly accessible and now includes more than 5.85 billion pairs of images and captions.
The images of children, the researchers found, came from mommy blogs and other personal blogs, maternity and parenting blogs, as well as snapshots from YouTube videos with few views, ostensibly uploaded for sharing with family and friends.
LAION spokesperson Nate Tylersays the organization has already taken action. “LAION-5B was taken down in response to a Stanford report that found links pointing to illegal content on the public web,” he says, adding that the organization is currently working with “theInternet Watch Foundation, the Canadian Center for Child Protection, Stanford, and Human Rights Watch to remove all known references to illegal content.”
In December, researchers at Stanford University discovered that data collected by LAION-5B contained child sexual abuse material.
See also: EU: Research in Meta on negative effects of Instagram, Facebook on children
The problem of deepfakes is even growing among students in US schools, where they are used to bully classmates, especially girls. Photos of children can be used to create CSAM, but the database could also reveal potentially sensitive information, such as locations or medical data.
"Children shouldn't live in fear that their photos might be stolen and used against them," says Hye.
LAION confirmed to Human Rights Watch that the images the researchers found existed and agreed to remove them. But Hye worries that what she found is just the beginning. It was a “small piece” of the data her team was examining, she says — less than 0.0001 percent of all the data on LAION-5B.
Last year, a German ad campaign used deepfake content to warn parents about the dangers of posting photos of their children online. But that doesn't address the issue of images that have already been published.
Hye says the responsibility for protecting children from this type of abuse falls primarily on governments and regulators.
🔒 Protect your privacy with Proton VPN
Swiss VPN from the creators of Proton Mail — strict no-logs policy, strong encryption, and built-in NetShield that blocks ads, trackers, & malware.
- ✔ No-logs, based in Switzerland (except 14-Eyes)
- ✔ NetShield: blocks ads, trackers & malicious domains
- ✔ Covers all devices — free version available
The link is an affiliate link — SecNews may receive a commission at no additional cost to you. It does not affect the independence of our article writing.
“I think children and their parents shouldn’t take responsibility for protecting children from a technology that they can’t protect themselves from,” Hye says. “It’s not their fault.”
See also: United Kingdom: Smartphone ban for children under 16?
Child protection
Unfortunately, since there are no effective protection measures, parents will have to take action to protect their children's images by avoiding posting them on social networks and other public platforms.
Additionally, by using privacy settings on platforms social media, parents can limit who can see and download their children's photos. It is important to regularly check these settings and adjust them accordingly.
It is also extremely important to inform children about the dangers of posting personal photos online. Educating children about online safety can help them understand the importance of protecting their privacy.
Finally, it is good practice to avoid using apps and tools that request access to photos and device files without a clear reason. Parents should carefully read terms of use and privacy policies before allowing access.
Source: www.wired.com
