AI is a force multiplier, not a substitute for human policing, warn safety experts

Rapid uptake of AI technology in the fight against crime should not obscure the need to improve human crimefighting capabilities, according to key safety & security stakeholders who spoke with the Cape Chamber this week. 

Crimefighting stakeholders say rapidly evolving AI requires a matching human response, as well as proper policy 'guardrails' to ensure the technology remains primarily an asset rather than a liability.  

Stakeholders shared their thoughts this week in response to Cape Chamber queries about recent technology-led crime prevention projects in the Cape metropole, including the Chamber's own tech-driven project in Philippi East.   

AI-backed surveillance and facial recognition has been deployed to good effect in Philippi East, with results of the pilot phase due to be released next month. However, the rollout coincides with concerns that AI evolution is outpacing human ability to manage the technology – particularly the potential pitfalls.  

"Is technology the solution or is technology a tool to enhance human capability?" Asks Dr Kelly Stone from the Institute for Security Studies Africa. "There's huge excitement about giving the police technology and less thinking around oversight and the capability of people to use it." 

"When it comes down to decision making, are we delegating too many tools as a way of subjugating our own powers, on the assumption  that technology is best?"  

The undoubted benefits of AI technology should not be a substitute for human intervention and response infrastructure. "Otherwise it's a case of knowing so much more about what crime is going on, but not changing anything about our ability to respond."   

Hubert Paulse, Cape Chamber Safety and Security Deputy Chairperson, said AI presented huge opportunity, provided it was properly deployed and managed. "Governance must keep pace with innovation.  It's the force multiplier and the value that comes from putting better information and capability into the hands of people responsible for making decisions and taking action," Paulse said.  

The debate comes as AI adoption accelerates across both public administration and criminal networks. While institutions leverage machine learning for fraud detection and risk profiling, syndicate operations are increasingly deploying generative AI for automated phishing, voice cloning, and deepfake scams targeting local consumers and businesses 

.