← Return to Insights

The Architecture of Trust: Governance in the Age of AI

• Governance & Compliance

We built StreetSignal on a premise of radical transparency. However transparency requires rules. In an era where “AI” often implies scraping and theft we take a different approach. We aggregate data from the City of Cape Town, Statistics South Africa and the South African Police Service to build a view of the market. This creates a responsibility.

The Data Sovereignty Doctrine

We do not scrape data. We do not infer trends from incomplete sets. We adhere strictly to the Spatial Data Infrastructure Act. This is how we ensure that when you see a Risk Signal it is backed by actuarial rigour rather than algorithmic guessing.

Our compliance framework rests on four pillars:

  1. City of Cape Town Open Data: We utilise the Open Data Policy (Policy Number 27781) which allows for commercial redistribution provided we acknowledge the source. We accept the “as is” warranty and process the data to remove anomalies before presentation.
  2. SAPS Crime Statistics: We source crime data via DataFirst at the University of Cape Town (UCT). This is licensed under Creative Commons Attribution 4.0. We cite the DOI (10.25828/5MAW-4H90) and comply with the National Treasury Secure Data Facility checking policy.
  3. Spatial Boundaries: The police station boundaries are protected by State Copyright. We do not redistribute the raw shapefiles. We map them internally to visualise risk zones without violating the redistribution clauses of the SDI Act.
  4. Privacy by Design: We aggregate data at the suburb level. We never display individual household data. This ensures full compliance with POPIA (Protection of Personal Information Act).

The “No-Black-Box” Promise

Institutional insurers do not trust black boxes. They trust audit trails. Every metric on StreetSignal can be traced back to a government gazette or an official statistical release. We do not use “AI” to hallucinate safety ratings. We use it to normalise disparate datasets into a unified schema.

If a suburb is rated “High Risk” it is because the raw crime density per capita exceeds the provincial median. It is a mathematical fact not an opinion.

Conclusion

Trust is our currency. By adhering to these strict governance standards we build a platform that is not just useful for home buyers but defensible in a boardroom. We are building the reference standard for African property intelligence and that starts with respecting the data.