Trust by design: Why smart cities need more than AI

By: TJ Hanekom, COO at Africonology

Artificial intelligence (AI) is everywhere in the smart-city conversation. We hear about AI-enabled cameras, facial recognition, predictive policing, intelligent traffic management and automated service delivery. AI undoubtedly has an important role to play. It can help cities see more, identify patterns, respond faster and make better decisions.

But having worked around the practical realities of connecting technology, data and people, I think we sometimes get ahead of ourselves. The question is not simply what AI can do. It is whether we can trust the information, identities and processes behind it.

Take the recent discussion around the planned rollout of AI-enabled body-worn cameras by SAPS. From an operational perspective, I immediately start asking different questions.

Who owns the footage? Who can access it? Where is it stored? How do we know it has not been altered? Can we prove when and where it was captured? Can it still be authenticated as evidence months or years later? And how do we establish that the person associated with a digital record is who they claim to be?

Those questions are not really about AI. They are about digital trust.

AI is only as good as the environment around it

AI rarely operates as a standalone technology. A camera captures information. A network transports it. A platform stores it. Another system analyses it. An operator acts on the result. The information may then be shared with another department or used as evidence.

There is a whole chain behind the AI. If an AI system identifies someone using poor-quality or incorrect data, a more sophisticated algorithm does not make that identification reliable. If a citizen applies for a municipal service online but their identity cannot be verified, how do we know the person making the request is who they claim to be? And if a traffic-management system is working from outdated or conflicting information, how useful is the intelligence it produces? These are practical challenges, not theoretical ones.

Four foundations of digital trust

For me, the smart-city conversation increasingly comes down to four things: trusted identity, trusted data, evidentiary integrity and governance.

First, trusted identity. As more services move online and multiple government systems interact, we need confidence that people are who they claim to be. If identity is compromised at the start of a process, everything that follows becomes questionable. Trusted identity therefore needs to be treated as foundational digital infrastructure, not simply a login mechanism.

Second, trusted data. Smart cities generate enormous volumes of information from cameras, sensors, vehicles, municipal systems and citizen interactions. But more data does not automatically mean better decisions.

We need to know where information came from, when it was collected, whether it has been changed and whether it is still relevant. I’ve often said that data is only useful when people trust it.

An algorithm can process information at extraordinary speed, but it cannot turn bad information into good information. If different departments have conflicting records about the same person, property or incident, adding AI does not necessarily solve the underlying problem. Often, the important work happens before the AI gets involved: connecting systems, cleaning information, establishing a reliable source of truth and ensuring the right people can access it.

Third is evidentiary integrity. The body-worn camera example brings this into sharp focus. If footage is transferred between systems, accessed by different people and stored for years, we need to be able to establish its provenance and demonstrate that it has not been altered.

The same applies to a traffic system recording an accident, a camera capturing an incident or a digital report informing an emergency response. Digital trust is therefore about more than cybersecurity. It is also about being able to prove that information remains authentic and trustworthy throughout its lifecycle.

Finally, there is governance. A city can have sophisticated analytics and highly integrated systems, but there still need to be clear rules about who can access information, what they can use it for, how long it should be retained and who is accountable when something goes wrong.

When AI influences decisions that affect people’s lives, the answer cannot simply be: “The AI said so.” There needs to be accountability around the technology and human oversight where it matters.

Integration is where the real work happens

Cities rarely start with a blank sheet of paper. They already have legacy databases, cloud platforms, cameras, access-control systems, geographic information systems, emergency-response platforms and departmental applications. The practical challenge is making this all work together.

A smart city is not created simply by buying another platform and adding another dashboard. If traffic management has one version of the information, public safety another and municipal services a third, the problem is not necessarily a lack of technology. It is a lack of integration and trusted information.

That is why the practical questions matter. What happens when a citizen moves house? When an identity record changes? When two systems disagree? When an officer needs to retrieve footage from an incident months later? When information needs to move between departments during an emergency? Those questions determine whether a smart-city solution works beyond the pilot stage.

Building trust into the smart city

The temptation with emerging technology is to focus on what is new: what can the latest AI model do, how much data can we process and how quickly can we automate a process?

But the more important operational questions are often less glamorous. Can we trust the identity? Can we trust the data? Can we prove the integrity of the evidence? Can our systems communicate? Can we explain how an important decision was reached? And is there clear accountability when something goes wrong?

If the answer is yes, AI becomes considerably more useful. If the answer is no, adding more technology can simply make an existing problem bigger and faster. The future of the smart city should therefore not be about pursuing AI for its own sake. It should be about creating the digital foundations that allow AI and other technologies to deliver meaningful, responsible outcomes.

Technology creates capability. Digital trust creates confidence in that capability. And ultimately, that is what smart cities need: not simply to be more connected or intelligent, but to be trusted.