Prototype: SOF IA – technology-facilitated Gender Based Violence (tfGBV)

Violence against women, in all its forms, exists as part of the ongoing patriarchal structure of dominance that prevails in our society. In the digital age, this takes the shape of technology-facilitated Gender-Based Violence (tfGBV), which has become an increasingly prevalent issue in our highly digitized world. As digital spaces have become an extension of our physical reality, the gender-based violence experienced by women in various contexts also finds expression through the use of technology and online platforms.

Despite being a persistent problem, Chile lacks comprehensive data and statistics on incidents of tfGBV. Moreover, there is a lack of policies in place to effectively measure, respond to, and provide support to women who endure such violations. The absence of a systematic approach to address tfGBV is not unique to Chile; it is a recurring issue in many countries. Although this problem was acknowledged over fifteen years ago, it has remained minimized, invisible, and unaddressed by the relevant authorities, whether political or law enforcement. Furthermore, there are a lack of institutional mechanisms to handle complaints and report cases that occur in women’s daily lives. In Latin America and other regions, feminist civil society organizations and collectives, including cyberfeminism initiatives, have taken the lead in providing support, response, guidance, and assistance in these cases. However, their efforts alone are insufficient to address the sheer quantity and diversity of cases that arise.The proposed system will operate on a web platform, serving as the initial prototype for the conversational agent. Following testing, it will be determined whether the conversational agent will be integrated with a messaging system or solely function through the web platform.

The design process will prioritize usability and user research (UX), drawing insights from simulated profiles or real-life cases. Before the project’s development phase, an initial set of focus groups were conducted with a group of women  actively engaged in feminist collectives and organizations who had previously experienced instances of tfGBV on their social networks. 

The envisioned user profile for the prototype is based on this user research: a woman who receives or undergoes tfGBV attacks on her social networks and is highly active within these platforms. This user profile includes opinion leaders, individuals who provide information on various social and political issues, such as human rights activists, communicators, journalists, and academics who influence public policy. 

To process the information from reported cases, an automated AI system will be employed, creating a database based on the reports. In the initial phase, this system will learn to distinguish various aspects, including i) types of attacks, ii) keywords associated with harassment/aggression/attacks, iii) profiles of aggressors or accounts involved, iv) frequency of cases, and v) recurrence of cases.

 A strategic objective of the project is to utilize AI to generate alerts when an aggressor profile is identified or when coordinated attacks or harassment against women occur on social media platforms, particularly within sociopolitical contexts of crisis or conflicts. The database resulting from the reported cases will facilitate the automated creation of data visualizations to be made accessible to the public. The development and programming will prioritize the use of open-source software, enabling the utilization of pre-existing resources for conversational agent programming and automation. The conversational agent  is intended to be launched at the end of September 2023.

Click here to watch the presentation