AI Training on Real Images of Children Raises Privacy Concerns

A recent report from Human Rights Watch has revealed that over 170 images and personal details of children from Brazil have been used to train AI without their consent. The images were taken from various online sources dating back to the mid-1990s, unbeknownst to the children or their parents.

The dataset in question, LAION-5B, is an open-source collection of data used to train AI models for startups. The images were originally gathered by Common Crawl, a data repository that scrapes the web for information. The dataset now contains links to billions of image-caption pairs, including those of children.

Hye Jung Han, a researcher at Human Rights Watch, expressed concerns about the violation of children’s privacy and the potential risks associated with AI-generated imagery. The images were sourced from personal blogs and videos, where the children were expected to have a level of privacy.

LAION, the organization behind the dataset, has taken steps to remove the flagged images and is collaborating with various entities to address the issue of illegal content. YouTube, where some of the images originated, has policies against unauthorized scraping of content and is working to combat such abuses.

The use of AI training data containing child sexual abuse material is a growing concern, as deepfake technology becomes more prevalent. Hye emphasized the need for government intervention to protect children and their families from potential harm.

While efforts are being made to address the issue, the challenge of removing such content from the web remains. The responsibility to safeguard children from these abuses falls on governments and regulators, as well as the organizations involved in AI development.

In conclusion, the incident underscores the importance of privacy protection and ethical considerations in AI development, especially when it involves vulnerable populations like children.

Hye Jung Han is a children’s rights and technology researcher at Human Rights Watch. She is dedicated to advocating for the protection of children’s privacy and safety in the digital age. With a background in human rights law, Hye has been actively involved in addressing the ethical implications of AI technology and its impact on vulnerable populations. Her work highlights the need for regulatory measures to ensure the responsible use of AI in society.