Predictive Policing and the Implications of AI

Predictive policing, driven by artificial intelligence (AI), is rapidly reshaping law enforcement strategies around the world. A recent in-depth legal analysis by Jacqueline Hahn examines the development, risks, and regulation of AI-powered crime forecasting in the United States, China, India, the European Union, and Germany, highlighting how different philosophies influence the balance between public safety and individual rights.

AI systems are now widely used to predict crime hotspots, allocate resources, and even generate risk profiles of individuals based on social, behavioural, and demographic data. These practices have achieved some success—such as significant reductions in gun violence in Richmond (US), or AI-assisted crowd monitoring in Delhi (IN)—but raise serious questions around bias, discrimination, transparency, and due process.

The note warns that AI systems often replicate historical injustices embedded in their training data, such as over-policing of minority communities in the United States or disproportionate surveillance in Muslim-majority neighbourhoods in India. China’s widespread AI surveillance is noted for prioritising state control over privacy, with systems like Skynet and Cloudwalk operating under minimal public scrutiny.

In contrast, the European Union emerges as the global benchmark for ethical and legal oversight. The GDPR and the newly adopted AI Act impose strict rules on data handling and prohibit certain high-risk applications, including real-time biometric surveillance in public spaces. Germany’s Constitutional Court has even struck down the use of predictive software like Palantir Gotham for violating citizens’ rights to informational self-determination.

While most countries lack binding AI legislation, the EU’s model represents the first serious attempt to balance innovation with democratic safeguards. The analysis concludes that future use of AI in policing should be governed by independent review, clear legal standards, and a commitment to transparency—elements largely absent outside the EU.

As law enforcement agencies integrate AI into operations, Hahn urges a shift from unregulated deployment to inclusive, legally anchored innovation, encouraging jurisdictions to adopt human rights-based models of technological governance.

🔗 Read the full article: Predictive Policing and the Implications of AI Across Global Frameworks – Columbia Law Review (Note by Jacqueline Hahn)