Breaking the Cycle: Unmasking the Flaws of Predictive Policing



Predictive policing software has gained significant attention in recent years for its ability to analyze crime data and make predictions about future criminal activity. However, researchers like Kristian Lum argue that these algorithms are merely a facade, hiding the flaws of a flawed system. In a talk by Lum, she highlighted the ineffective nature of using arrest data to guide future police searches and arrests. Lum’s argument raises questions about the root causes of crime and the need for a more comprehensive approach to addressing these issues.

The Feedback Loop of Bad Policing:

Lum’s point centers around the idea that the current approach to policing is repetitive and misses the mark when it comes to tackling the root causes of crime. She highlights how, in drug crime cases, arrests are made based on past arrest data rather than a comprehensive exploration of the entire area. The result is a cycle where the same neighborhoods and demographics are targeted repeatedly, even though drug crime is prevalent across all areas.

Predictive Policing Software: A Cover-Up for Ineffective Policing:

Predictive policing software, according to Lum, does not offer truly novel solutions but instead perpetuates existing practices. The training data used in these algorithms is derived from the flawed arrest data, which further entrenches the patterns of bad policing. The black-box nature of machine learning algorithms in predictive policing software conceals the underlying issues and creates a misleading perception of efficiency in law enforcement.

Understanding the Root Causes of Crime:

To make a meaningful difference in crime reduction, Lum argues that it is essential to examine why crime occurs in the first place. She asserts that most people commit crimes out of necessity, either due to extreme poverty or a lack of access to resources. By addressing these underlying issues and providing support to impoverished communities, there is a greater likelihood of making a lasting impact on crime rates.

The Complexity of Crime and Criminology:

Beyond poverty, Lum acknowledges that certain crimes may stem from anti-social personality disorders or political motivations. However, she argues that abstract models that attempt to explain crime as an epidemic or a consequence of broken social norms are oversimplifications of a far more complicated issue. By recognizing the diverse factors that contribute to criminal behavior, society can create targeted interventions that address the root causes instead of just managing the symptoms.

The Need for an Integrated Approach:

Lum’s perspective underscores the importance of adopting an integrated approach to crime reduction. This involves not only addressing poverty and socioeconomic disparities but also investing in mental health support, education, and rehabilitation programs. By taking a comprehensive view of the factors influencing crime, societies can work towards creating a safer and more equitable environment for all.


Kristian Lum’s critique of using predictive policing software highlights a significant flaw in the current approach to law enforcement. It calls for a shift in focus towards understanding the root causes of crime and implementing measures that address these underlying issues. To make a meaningful impact on crime reduction, a multifaceted approach that tackles poverty, invests in mental health support, and promotes equal opportunities is essential. Only then can society move towards a more just and secure future for all.

Disclaimer: Don’t take anything on this website seriously. This website is a sandbox for generated content and experimenting with bots. Content may contain errors and untruths.