Published on 23 September 2023.

Introduction:

Predictive policing, a concept that employs computer algorithms to forecast potential criminal activities, has garnered significant attention in recent years. While proponents argue it can enhance law enforcement efforts, there are growing concerns about its fairness and effectiveness. Recent reports from New Orleans shed light on the complexities of implementing predictive policing and highlight the need for transparency, oversight, and community involvement.

Flaws as Seen from New Orleans

1. Lack of Transparency and Community Support: The New Orleans experience underscores the importance of transparency and community involvement in adopting predictive policing. The program, initiated by the data mining company Palantir, was introduced without public input or approval from elected representatives. To address this, initiatives like Community Control Over Police Surveillance (CCOPS) are essential, ensuring that communities have a say in the deployment of surveillance technologies.
2. Flawed Data and Racial Bias: Predictive policing relies heavily on data, but when this data is flawed or tainted with bias, it can perpetuate injustice. In New Orleans, data sources included field interview cards (FICs), which are known to have racial biases. Similarly, arrest and conviction data often have racial disparities. Predictive algorithms, when fed with biased data, produce biased results, exacerbating racial disparities in law enforcement.
3. Carrot and Stick Approach: The New Orleans program adopted a “carrot and stick” approach, promising social services and support to individuals identified as at-risk. However, over time, the emphasis shifted towards punitive measures. This raises concerns about whether predictive policing can genuinely provide benefits or merely lead to increased surveillance and prejudicial encounters with law enforcement.
4. Social Media and Privacy Concerns: While the New Orleans program denied the bulk collection of social media data, the use of social media in predictive policing remains a contentious issue. The potential for chilling free speech, inhibiting organization, and threatening privacy is significant. Clear guidelines and safeguards are needed to address these concerns.
5. Unproven Effectiveness: The termination of New Orleans’ contract with Palantir raises questions about the effectiveness of predictive policing. While proponents tout its potential, the limited success and termination of contracts suggest that these technologies may not live up to their promises.

Conclusion:

Predictive policing is a complex and evolving field, and the New Orleans experience offers valuable lessons. Transparency, community involvement, addressing racial biases in data, and a careful balance between rewards and punishments are crucial to ensuring that predictive policing benefits society without compromising civil liberties. As discussions on predictive policing continue, it is imperative that these lessons guide its responsible implementation in the future.