Facial recognition technology is increasingly being integrated into CCTV systems across South Africa, but many people remain unaware of its presence. In some shopping malls, for instance, CCTV facial recognition is already in use, yet there is often no signage indicating this, raising concerns about transparency and compliance.
When Solution House began exploring facial recognition technology for CCTV systems four years ago, it quickly identified this lack of clarity as a significant problem. "What is happening and why is this a problem? People don’t know what they may or may not do under POPIA," explains Tiaan Janse van Rensburg, Technology Director at Solution House. "The financial cost of getting legal advice is massive, and the entire POPIA framework – especially concerning physical security systems – can feel overwhelming for businesses."
The uncertainty around POPIA compliance has left many organisations hesitant to fully adopt facial recognition technology. Some choose not to implement it at all, while others proceed discreetly, fearing potential legal repercussions. This hesitation often stems from a lack of understanding of the law and the technology itself. For those who do move forward, selecting the right technology is crucial. A successful facial recognition system requires a high-quality matching platform, capable of accurately identifying individuals as they pass by a camera. However, as Janse van Rensburg notes: "Not all matching platforms are the same. Many businesses rely on the systems provided by their CCTV manufacturers, but these companies are not always specialists in facial recognition."
To address this, Solution House developed the FaceCamAlert solution and invested in specialised matching platforms that deliver superior accuracy, though this approach can be more expensive than using the default options. "The investment is necessary to ensure the technology works effectively," says Janse van Rensburg. However, the technology alone is not enough. Legal compliance adds another layer of complexity, particularly when managing data across multiple locations.
For example, a shopping mall may maintain a list of individuals who have been caught shoplifting. If someone on this list enters the mall again, a well-designed matching platform can quickly identify them. However, if that person shoplifted at a different mall, the system might not recognise them due to the lack of communication between different properties. This gap is largely due to the challenges of POPIA compliance and the technical difficulties in linking these environments.
To overcome this, Solution House has developed solutions that can manage multiple environments simultaneously, a process known as multi-tenancy. "Multi-tenancy is the ability to combine numerous elements into one cohesive solution, and that’s what we specialise in," says Janse van Rensburg.
Achieving POPIA compliance is a cornerstone of Solution House’s approach. Despite the fact that POPIA does not specifically address facial recognition, Solution House has invested heavily in building a compliant solution. "We did a massive data assessment and created a legal framework to determine where our facial recognition solution fits within existing laws," Janse van Rensburg explains. “With this foundation, we have developed a management framework that aligns with POPIA’s broad regulations.”
The framework also includes severity levels that determine how long data is retained and who can access it. For example, data related to less severe incidents is kept for a shorter period and shared with fewer people. In cases where someone is caught shoplifting, the process is more complex. After signing an admission of guilt, the individual should also sign a further processing notice, allowing the data to be shared with third parties. Furthermore, depending on the data being shared, the responsible party may have to apply to the regulator to ensure full compliance.
Solution House goes beyond providing the technology. It assists clients in navigating the legal landscape by offering management frameworks, signage and legal opinions, ensuring that its systems are fully compliant with POPIA. "We take POPIA extremely seriously," emphasises Janse van Rensburg. "Our FaceCamAlert solution currently focuses on private property, where individuals have the right not to be tracked without proper cause. We use facial recognition matching in support of these laws, and we don’t store unnecessary data."
Instead, Solution House operates with watchlists compiled within its management framework. If there’s no match, the data is deleted immediately. If there is a match, specific actions are triggered based on the situation. Individuals have the right to inquire about their data, request its deletion and challenge its accuracy. However, if they are on a watchlist, the situation may be more complex, as there are legitimate reasons for their inclusion.
Through rigorous legal preparation, investment in top-tier technology and a commitment to ethical data management, Solution House offers businesses a way to leverage facial recognition technology while remaining fully compliant with POPIA. Their approach not only solves the technical challenges but also ensures that the legal and ethical dimensions of facial recognition are fully addressed.
You can read more at https://facecamalert.co.za/ or https://www.myincidentdesk.com/.
Share