Top 93%
in this catalogue (794 scored practices)
Scores cluster high, so position within the catalogue is often more telling than the number alone.
Evidence of impact / measured public value1/3
City-reported alert and arrest figures exist (3,893 alerts, 50+ arrests, 35 firearms recovered in 2024/25) but come solely from the city and its supplier, with no independent verification of impact on gun violence.
Transparency, fairness & accountability0/3
A 2026 academic review found no independent public audit had ever assessed the system's accuracy or false-alert rate, unlike US jurisdictions with inspector-general oversight.
Transferability / demonstrated replication2/3
ShotSpotter/SoundThinking technology is deployed in dozens of cities worldwide, demonstrating clear technical replicability, though Chicago's parallel deployment was cancelled in 2024 after producing evidence of a gun crime in fewer than one in ten alerts.
Scalability beyond pilot1/3
Coverage expanded from two pilot neighbourhoods in 2016 to several more apartheid-era settlements, but the city ended the three-year tender in mid-2026 and placed the programme under review rather than scaling it further.
Governance, capability & sustainability0/3
The programme ran on a renewable city tender with no independent oversight body, and critics noted it was deployed in poor communities with little resident consultation before the council let the contract lapse.
Evidoria. Cape Town's ShotSpotter Gunshot-Detection Network — A Cautionary Case in Unaudited Policing AI. Persistent ID: 0c5840ce-e599-4d05-b4c8-05610f7fe105.