LAHORE: The provincial capital police have adopted an artificial intelligence (AI)-based crime prediction system to identify potential crime hotspots and forecast patterns of criminal activity before they occur.

The system, designed by Punjab Safe Cities Authority (PSCA), is being hailed as a major step toward predictive policing in Pakistan. It will use three years of recorded FIR data, Safe City camera feeds, and real-time field intelligence to help allocate resources more effectively across the city.

The police officials say the system will act as a “strategic multiplier,” enhancing existing policing methods that have already yielded historic success in reducing crime.

An official said that a high profile meeting held to review the AI-based crime prediction system here the other day analysed the official data compiled by the Lahore police operations wing. According to the figures presented in the meeting, he said, the registered property crimes dropped by nearly 77 percent between 2023 and 2025, reflecting one of the sharpest declines in urban crimes across the country. In 2023, 80,827 property-related cases were registered, which fell to 45,865 in 2024 and went further down to just 18,558 cases by September 2025.

PSCA designed system uses three years of FIR data, Safe City feeds and real-time field intelligence to help allocate resources more effectively

The meeting was told that the trend was consistent and city-wide, underscoring the effectiveness of Lahore Police’s targeted operations, community engagement, and enhanced accountability mechanisms.

A closer look at the September 2025 statistics showed a dramatic improvement compared to previous years. Total reported crimes decreased from 6,941 cases in September 2023 to 3,146 in 2024 and just 1,279 in 2025, marking an overall reduction of 82 percent in two years.

Robberies dropped by 91 percent, snatching by 89 percent, motorcycle theft by 74 percent, and burglary by 63 percent. Even serious crimes such as dacoity and robbery-cum-murder declined by 50 and 67 percent, respectively.

Similarly, analysis of the emergency 15 calls data showed a clear downward trajectory. Crime-related emergency calls fell from 76,710 in the first nine months of 2023 to 51,352 in 2024 and further down to 34,360 in 2025, a 55 percent decline in two years.

Robbery-related calls dropped by 66 percent, snatching by 61 percent, and motorcycle thefts by 53 percent, highlighting both a reduction in incidents and a rise in citizen trust.

The official said that DIG Operations Faisal Kamran told the meeting that the new AI system would build upon this foundation of success, allowing Lahore police to shift from reactive to preventive policing.

He said, “We are using artificial intelligence to anticipate criminal trends and identify vulnerable areas before crimes

happen. This innovation, brought in line with the vision of Punjab Chief Minister Maryam Nawaz and Inspector General of Police Dr Usman Anwar, will complement our on-ground intelligence, community policing, and patrol response systems that have already proven effective in reducing crime across Lahore.”

The DIG credited the continuous decline in crime to a combination of data-driven operations, Safe City surveillance integration, divisional accountability, and field intelligence-led policing. He also noted that consistent crackdowns on repeat offenders, improved resource deployment, and better citizen coordination had collectively helped make the city more secure.

Police officials emphasised that the AI initiative would not replace human policing but was aimed at enhancing decision-making and response precision, ensuring that police presence was concentrated where and when it was most needed.

The model is currently being calibrated through historical FIR data and will soon provide predictive maps to support patrol planning and rapid deployment.

Published in Dawn, October 6th, 2025