Harnessing Artificial Intelligence (AI) to Protect Latin American Children Online
"Being safe online means being able to have a healthy and safe development in the digital context, being able to learn and to thrive, without the fear of being bullied or being abused by others."
-Lina María Saldarriaga Mesa, Director of Operations and Project Lead, Aulas en Paz
Harnessing Artificial Intelligence (AI) to Protect Latin American Children Online
In an exciting collaboration, Universidad de Los Andes, alongside Programa Aulas en Paz, is furthering the fight against online child sexual exploitation and abuse in Latin America. Thanks to the support of the Tech Coalition Safe Online Research Fund, this initiative aims not only to understand the intricacies of online abuse but also to equip authorities, industries, and families with tangible tools to prevent this heinous crime.
In the Latin American context, this project is particularly relevant due to the increasing number of children accessing the internet, yet there remains a lack of evidence and insufficient research on the prevalence and impact of child sexual exploitation and abuse (CSEA) in the region. Moreover, legislation, investigation, and prevention measures have only recently begun to be implemented across these countries.
The Catalyst: Cross-Sectoral Collaboration Empowers Spanish-Language AI Solutions
At the heart of this crucial initiative is the pivotal investment from the Tech Coalition Safe Online Research Fund in independent research in languages other than English. This funding empowers Universidad de Los Andes and Programa Aulas en Paz to pioneer innovative solutions utilising artificial intelligence, as well as address the urgent need for protecting analysts and content moderators in the Spanish-speaking world. As one of the first Spanish-language tools of its kind, this project fills a critical gap in the realm of online safety, where limited efforts have been made to design and deploy research and solutions fully grounded outside of English-speaking contexts. By converting language into numerical data, the AI tool not only facilitates the identification of potential offenders but also shields analysts from direct exposure to harmful content, safeguarding their mental well-being. "The artificial intelligence tools not only process the information reports faster but also allow the analyst to be less exposed to the harmful content," adds Lina María Saldarriaga Mesa, one of the project leads.
Moreover, this endeavour exemplifies the power of cross-sector collaboration, with close coordination between different groups to develop the AI toolkits. "To create our tools, we work with engineers, psychologists, and lawyers, and even law enforcement," says Lina María Saldarriaga Mesa, and she adds: "It was through that collaborative work that made it possible for the artificial intelligence models to work correctly." The partnership with Los Andes and Red PaPaz through the Te Protejo hotline (a hotline which provides a platform for reporting crimes such as grooming, online bullying, and sexual extortion related to child sexual exploitation and abuse), both Safe Online grantees, demonstrates how Safe Online is facilitating such strategic collaboration to enhance online safety efforts. Additionally, the project uniquely combines the evidence generated by the AI tool linked to Te Protejo with insights revealed by children and parents through direct interviews, providing a comprehensive understanding of online threats and enhancing protective measures.
This intersection of cutting-edge technology and direct community engagement not only underscores its relevance and utility to the tech industry but also sets an example for future collaborations in online safety research and development. "Being able to transfer these tools is the key to sustainability for us. Both civil society and authorities are starting to see the benefits of the tools that we have created for them," shares Lina María Saldarriaga Mesa.
Understanding and Mitigating the Threat
The project has two overarching goals: firstly, to analyse and understand the language and behavioural patterns of online predators, and secondly, to develop and evaluate mitigation strategies. By leveraging advanced AI tools, researchers decode the intricacies of communication between potential aggressors and their victims. This insight forms the basis for the development of adaptable AI tools designed to be deployed by law enforcement agencies across the region. It is significant not only to get a better grasp of the issue but also to proactively seek solutions. The project team emphasises that their goal is not merely to understand the dark corners of online child exploitation but to shine a light on them through prevention efforts.
Strategies for a Safer Digital World
Parental mediation and industry-created protection tools are in the spotlight as potential safeguards against the growing threat. The study aims to further explore and evaluate the effectiveness of these strategies, offering a comprehensive understanding of how families and industries can actively contribute to child safety in the digital realm.
Designing Adaptable Solutions
The project team highlights two crucial factors in establishing a safer online environment for children and adolescents: firstly, the development of practical tools that not only align with academic research but also effectively support the work of law enforcement and authorities; and secondly, the necessity of collaboration across disciplines, including engineers, psychologists, lawyers, and law enforcement, which was instrumental in ensuring the functionality of the artificial intelligence models. The ultimate goal of the project is to establish AI tools that can adapt to diverse contexts within the Latin American region. The adaptability factor is crucial, given the socio-cultural diversity across countries. The tools developed do not only aid law enforcement agencies but are also valuable assets for families, industries, and governments in their collective mission to shield children from online harm. Lina María Saldarriaga Mesa states: "We are very excited by the fact that these authorities and hotlines are telling us that the tools that we have developed are both applicable and useful for their work." These applications are set to redefine the way authorities and organisations approach online child protection, ensuring a safer digital space for the region's youth.
A Latin American Initiative Looking Ahead for a Safer Digital Future
This project holds immense significance in the Latin American context, where the fight against online child exploitation is evolving rapidly. By not only understanding but actively preventing this crime, Universidad de Los Andes and Aulas en Paz aim to set a precedent for collaborative efforts that transcend borders, ensuring the safety of children online becomes a shared responsibility. Looking ahead, the team is optimistic about the transformative impact their research can have. In a world where the digital landscape is fraught with dangers for some of the most vulnerable members of society, initiatives like the one initiated by Universidad de Los Andes serve as a milestone of optimism, demonstrating that with the right tools and support, a safer online future is within reach.