Scarface predictive policing deployed at mines
Information Certainty: Documented
Deployment Purpose: Surveillance
| Summary |
|---|
| Bidvest Protea Coin, a South African security company, has deployed a high grade system for the tracking of suspect persons at mines around the country. The system is termed 'Scarface' and consists of an algorithm for identifying 'early' signs of transgression. It uses facial recognition and appears to be connected to other criminal databases. It can be described as predictive policing to the extent that it makes a claim to the early identification of the 'signs' of suspicious individuals, aiming to prevent crime before it occurs. It also claims not to rely on the intervention of a person (user) to make an analysis. It appears the system may not be limited only to mines and may serve as the bedrock for a general predictive policing system. |
Products and Institutions:
| Product Deployed | Bidvest Protea Coin Scarface |
|---|---|
| Institutions ⠉ | Bidvest Protea Coin |
| Datasets | Unknown Dataset 0160 |
| Search software |
Status and Events:
| Status | Ongoing |
|---|---|
| Events | Start (21 June 2022, Documented, , No description) |
| Start Date | |
| End Date |
Users:
| Involved Entities | Unknown Institution 0092 |
|---|---|
| Managed by | Bidvest Protea Coin |
| Used by |
Location:
| City | Pretoria |
|---|---|
| Country ⠉ | South Africa |
Description[ ]
A security company has deployed a high grade system to predict and track suspect individuals at mines around South Africa.
The Bidvest Protea Coin group has provided further details about the advanced facial recognition technology it’s using to track and catch criminals in South Africa – a system known as ‘Scarface’. Scarface is a platform that uses advanced Artificial Intelligence (AI) algorithms to help security teams identify what it refers to as ‘persons of interest’. The group said that the technology was originally developed as an early warning system that could identify potential threats and prevent assets or criminal attacks before they happened. “Since we know that crime often involves scouting, which takes place over a period, Scarface stops that process at the early stages before it matures with detailed and often classified information, by allowing early detection and/or apprehension of transgressors. Overall the system provides increased risk mitigation through a proactive approach to suspect identification.” The technology is fully autonomous and requires no user intervention to function as a recognition platform 1
But in order to work as intended, it needs access to a database of biometric information. An accurate database. It’s already got one, of a sort. It’s used to detect the presence of known illegal miners, or zama zamas, in some areas of South Africa. This, coupled with high-resolution cameras and the autonomous Scarface system, lets the group identify if there’s a ‘person of interest’ in the area. The system is apparently quick, and accurate — though accuracy rates aren’t mentioned. Still, it’s great… until it isn’t. The company doesn’t explain how it will deal with false positives, where it gets its facial data from, how broadly the program will roll out, and how the data collected will be managed. Because, yes, non-criminal facial data will probably be collected by, or be resident in, Project Scarface 2
References
- a b "This South African security company is using a tool called ‘Scarface’ to track people – how it works". (2022) <https://businesstech.co.za/news/technology/598848/this-south-african-security-company-is-using-a-tool-called-scarface-to-track-people-how-it-works/>
- a b "Scarface is watching you - How the South African facial recognition system works - StuffSA". (2022) <https://stuff.co.za/2022/06/22/scarface-is-watching-you-how-the-south-african-facial-recognition-system-works/> Accessed: 2022-12-06