Skip to content
An official website of the OECD. Find out more
Created by the Public Governance Directorate

This website was created by the OECD Observatory of Public Sector Innovation (OPSI), part of the OECD Public Governance Directorate (GOV).

How to validate authenticity

Validation that this is an official OECD website can be found on the Innovative Government page of the corporate OECD website.

Integrated System for Attention of Gender-violence Victims (using Artificial Intelligence)

shutterstock_1975202012-685x368

The innovation was born amid the pandemic and was inspired by the thousands of women that were locked up and forced to live with their aggressors. Just as an example, in 2020, the 144 hotline received >108 thousand communications related to domestic violence. We developed a comprehensive system of care for victims of gender violence in order to improve the state's response to the problem. The innovation consisted not only the gathering and systematizing of information, but also considered assessing the risk of femicide in each case through the use of artificial intelligence.

Innovation Summary

Innovation Overview

The State is late in reaching out to victims of gender violence and preventing femicides. In Argentina, during the year 2021 a femicide occurred every 29 hours, and until July 2022 there were 171 femicides and 3 transvesticides. In 2022, 191 daughters and sons were left motherless, 54% of the aggressors were partners or ex-partners of the victims and 60% were killed in shared housing.

The solution consists of an Integrated System of Care for Gender Victims with Artificial Intelligence (SIAVIGia). The objective of the platform is to provide governments with a tool that allows them to comprehensively manage the attention to victims of gender violence, and to have a common Anonymous Data Registry (ADR) in order to produce statistically sound evidence and information that supports the design of public policies related to this issue.

The innovation integrates a Risk Assessment system with AI that allows to estimate the probability that each domestic violence case escalates to greater conflict and even feminicide. The evaluation process allows to input several risk factors present together with an intensity measure to, as next step, weight these variables and assign scores, which allow to identify and prevent possible feminicide cases. SIAVIGia is a design and data-based experience. It pulls out data from all beneficiaries of social development and gender programs, integrating some working streams and making its management more efficient. It displays modules clustered in seven areas: 1) Social Security, 2) Citizen Security, 3) Health, 4) Diversity and Gender, 5) Justice, 6) Attention to Victims, and 7) Social Development.

The platform has enabled and protocolized channels to receive reports on cases of gender-based violence from any government agency, which can be integrated with any communication channel. The solution improves the institutional response by creating new reporting mechanisms, centralizing and systematizing information, coordinating and communicating the actions and policies of the agencies involved, and defining a Common Alert Protocol for victim care. The objective of the innovation is to make public management of gender issues more efficient, to streamline the state response in cases of violence, and to detect and prevent femicides.

Anyone who is a victim of gender-based violence, or who requires state assistance for gender-related reasons can benefit from the innovation. As gender violence is understood as a broad problem, not limited only to adult women, we target as the beneficiary population any person regardless of gender, sex or age. Socio-economic factors are taken into account as part of the risk assessment, meaning that populations in situations of poverty and vulnerability would be particularly visible in our model. This condition of vulnerability determines the prioritization in the attention of this populations when it displays a greater risk to physical integrity. Furthermore, the Risk Assessment considers the different factors of gender and diversity, addressing the problem of gender violence with a multisectoral approach, but deploying a centralized system for collecting and evaluating information.

The innovation is a useful, scalable and replicable product so that local governments can have a better tool to address this problem, therefore, the product can be offered in different formats, in particular as software as system.FSIA accompanies the implementation through assistance, protocolization of actions, and training of government personnel. It also incorporates tools for public management. This can be financed directly by the municipality or can be transferred thanks to funds that subsidize it to the end client.

The main scalability model of the solution is replication making use of the different public channels. In a broad sense, any government that addresses gender issues is an organization that can implement the solution. Most governments do not have systems of this type, let alone Risk Assessments independent of human agents. The Municipality of Lomas de Zamora was chosen for the implementation of the first pilot because of its agenda highly devoted to innovation in public management and its vocation to address important issues with innovative technologies.

Innovation Description

What Makes Your Project Innovative?

The innovation lies in the development of a Femicide Risk Assessment: a system that uses Artificial Intelligence and learns to evaluate the risk of femicide in a case of gender violence, for which it assigns different levels of priority in the follow-up of each case. According to reports from EIGE (EU), no such tools have been successfully implemented in Latin America prior to the development of this innovation. Previously, the determination of the risk of femicide present in a case of gender-based violence depended exclusively on field experience and the intuition of the police agents who attend and intervene in the cases. In that way, the innovation allows to calculate the presence of risk factors, the intensity with which they are present, and their abstract priority by incorporating algorithmic weighting.

Innovation Development

Collaborations & Partnerships

The main ally was the Municipality of Lomas de Zamora as main implementing partner. In particular the Gender Unit. Their support was critical in enabling the different decisions and procedures to implement the solution as a single registry and the incorporation of AI. Also the British Embassy, as they financed the development of the solution.

Users, Stakeholders & Beneficiaries

The main users are the officials from the different areas of the Municipality of Lomas de Zamora, in particular those directly working with gender, health, security and related policies. Relevant to mention is the neighborhood and local NGOs that supported the data analysis. Moreover, the beneficiary population is all victims of gender violence who obtain better care and are not re-victimized in each interaction with the public authorities.

Innovation Reflections

Results, Outcomes & Impacts

The impact of the solution is measured by evaluating the differences between the level of reporting and the levels of response to gender issues. This evaluation also considers the evolution of gender-based violence cases, and the evolution of feminicides within the territory.

Between December 2019 and November 2020, a total of 1147 cases of gender-based violence were admitted. They triggered a total of 3926 interventions on site, which represented a total of 3.5 interventions per person during that period of time. For the purposes of impact measurement, it is important to consider both qualitative and quantitative indicators that monitor a broad spectrum of variables related to the multidimensional approach to the problem.

Challenges and Failures

The greatest risk for this implementation is political and its interplay with technology. The Foundation aims at implementing technology in a context where HR can be a bottle neck for any other reason. FSIA mitigates these risks through training to ensure effective implementation. Another challenge we encounter is the lack of organizational culture of information sharing.

As local agencies work in isolation from other government agencies, we decided to co-construct a multidimensional approach to victim care which was able to better understand the comprehensiveness and singularity of each case, to coordinate actions and to work collaboratively.

Conditions for Success

Political will is key to achieve success in this kind of policies. Technological solutions for governments require political consensus for their implementation, since the specific areas that must adopt the tools must have the will to incorporate them into their work routine and make them their own for themselves and their team.

In this sense, it is essential to create conditions in which the HR are exposed to training instances to become familiar with the tool and its functionalities for the purpose of a gradual incorporation into the work methodology. Over time these new behaviors are expected to become a new organizational culture based on communication and understanding of the situation of every victim.

Replication

SIAVIGia is a useful, scalable and replicable innovation that local governments can use for social, security and gender issues to address different problems. During 2022 the Municipality of "Almirante Brown" decided to implement the system.

Lessons Learned

We learned about data organization and about the relevance of sharing data among management areas to optimize the State response. We also learned to manage more complex and dynamic information platforms to enrich the volume and quantity of the available data for analysis and policymaking. On top of this, we were able to learn from the importance of promoting inter-agency cooperation to prevent domestic violence. For example, public health assistance services can also provide relevant information about the nature of specific injuries or conditions related to domestic violence.

Year: 2020
Level of Government: Local government

Status:

  • Implementation - making the innovation happen

Innovation provided by:

Media:

Date Published:

16 November 2022

Join our community:

It only takes a few minutes to complete the form and share your project.