The Algorithm Declared Her Safe, Then Her Husband Took Her Life
Spain has been using the ‘VioGén’ algorithms to assess the likelihood of repeat attacks on domestic violence victims and determine the level of protection they should receive.
On January 11, 2022, Lobna Hemid was attacked by her husband, Bouthaer el Banaisati, in a small apartment outside Madrid using a fragment of a shoe rack. A clear as day expamle on how sometimes increase technological reliance has had tragic mortality rate consequences.
Despite Lobna producing photos of swellings on the body the indicated frequent abuse by her husband, coupled with a police report of el Banaisati violent tendencies, VioGén still classified her risk as low.
Seven weeks later, Lobna was violently stabbed and murdered by her husband, before he turned the knife on himself and killed himself.
In this case, the algorithmic system, VioGén, played a key role in unravelling of Lubna’s death and her husband’s brutal attack. The enforcement Spanish system, which provides the police with provision of services, such as genetic analysis of evidence left by the culprits, failed to see and register the attacks as violent acts of domestic abuse.
VioGén has thus turned out to be more like a standard asset in the policing structure in Spain. It is mentioned that one cannot differentiate where the human input ends and where the algorithm comes into play.
The system has successfully caused a low frequency of repeat attacks in domestic violence cases, but at the same time such cases were also misclassified by the system itself, subjecting the victims to further abuse and, in some instances, death. Nowadays, VioGén is applied to assess 92,000 ongoing feminicides in Spain. Yet 8% of the women classified as low risk/ negligible risk said they were harmed once more, and 14% those classified as low risk.
Impact on Protection Measures
Spanish police are taught to override VioGén on any decision with sufficient cause but said to say yes to 95% of the algorithm’s risk scores, which are also applied by the judges in cases of restraining orders and other forms of protection.
Pavlou Susana, a director of the Mediterranean Institute of Gender Studies added that the algorithm dismisses the police from the responsibility to evaluate the situation and decide what the victim requires.She mentioned that the algorithm “kind of absolves the police of any responsibility of assessing the situation and what the victim may need.”
How VioGén Works
VioGén was designed for police to focus on the specific cases and assign resources based on the women vulnerable to further abuse. Again, created with consultation with statisticians, psychologists, and other professionals, the system employs a questionnaire that is completed by police after taking a statement from the victim. The responses are then matched against historical data and a risk capital score varying from minimal to severe is given. More attention and service are provided to those who are recognized as high-risk victims through patrol and shelter protection and so on while less service is provided to those who have lower scores.
Even in this case, however, accuracy of the VioGén precisely correlates with the quality of the input signals. Since the victims fear their offenders, are ashamed, and in certain cases are economically dependent on them, they may hide the true reality they are living in from the public’s eye. This causes false risk scores.
Employed police officers which can be busy most of the time, may not take their time to study every case, which degrades the algorithms authenticity even more. Professionals in the field and even agencies for victims’ rights prefer statements that exclude police officers as the primary investigators of such occurrences. They also retain that suspects should not stop victims from taking along someone whom they trust when being interrogated to unveil more of their ordeal.
Spanish officials are expected work and experiment on integrating more sophisticated stages of AI in VioGén by tapping from another data source. Juan José López Ossorio, one of the VioGén founders, has emphasized that optimality will never be achieved while stressing the importance of its constant improvement as a way to enhance the degree of victims’ protection.
Final Thoughts
While VioGén has been effective in reducing overall violence against women in Spain, its limitations highlight the importance of continuous improvement and human oversight. As Spain navigates the complexities of integrating AI into domestic violence prevention, it must ensure that technology enhances, rather than replaces, thorough and compassionate human intervention.
Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Intelligent Tech sections to stay informed and up-to-date with our daily articles.