Imported gunshot detection systems can marginalise communities

Surveillance technologies imported from the Global North into postcolonial cities can reshape public debate, governance and understanding of safety in ways that deepen inequality and marginalise community voices.

This is according to a research paper co-authored by the University of Cape Town’s (UCT) School of Information Technology’s Grant Oosterwyk. The paper examined the rollout of ShotSpotter – a United States-developed acoustic gunshot detection system, in Cape Town. It explored how surveillance technologies are legitimised in cities shaped by apartheid spatial legacies, institutional mistrust and structural inequality.

“Our paper showed that when technologies such as ShotSpotter travel from the Global North into post-colonial contexts, they often arrive, not simply as tools, but as governance scripts that shape how problems are understood.”

“Our paper showed that when technologies such as ShotSpotter travel from the Global North into post-colonial contexts, they often arrive, not simply as tools, but as governance scripts that shape how problems are understood and which solutions appear legitimate,” he said.

A detailed analysis

Using a Habermasian critical discourse analysis, Oosterwyk and his colleague Raffaele Ciriello of the University of Sydney, analysed parliamentary debates, municipal communication, media reporting, civil society documents and community testimonies related to ShotSpotter’s rollout in Cape Town. Their research findings indicate that imported surveillance systems are often presented as neutral, data-driven solutions to violence, while obscuring deeper questions about democratic accountability, commercial interests and community participation.

He said the research study identified four recurring, discursive strategies used by officials, vendors and political actors to frame ShotSpotter as a necessary and authoritative policing intervention. These, he said, include technical and definitive language that presents the system as objective and unquestionable; numerical claims presented without independent auditing; passive constructions that obscure vendor responsibility; and metaphors that frame policing as warfare or technological combat.

Oosterwyk explained that these narratives create a “techno-solutionist” understanding of violence, where deeply rooted social and political challenges are reduced to technical indicators such as response times, detection accuracy and arrest statistics. The paper also highlighted that the same promotional language used around ShotSpotter in the US travels to South Africa, but without the same accountability infrastructure.

“Contexts marked by historical inequality and institutional instability, these imported scripts risk misrepresenting harm, marginalising local knowledge and narrowing democratic deliberation at the very moment [that] inclusive governance is most needed,” he said.

Who defines the problem?

He said in several US cities ShotSpotter has also faced independent audits, legal challenges and public oversight.

“In Cape Town similar claims about accuracy and effectiveness have circulated for years without equivalent mechanisms for public scrutiny. That difference matters because it changes the democratic weight those claims carry,” Oosterwyk said.

And surveillance governance should not be evaluated through technical performance metrics, but also through questions of legitimacy, transparency and public participation. 

“Technical audits alone cannot capture the political and institutional dynamics that shape the real-world effects of dragnet surveillance. We need to ask who defines the problem, who controls the evaluation criteria and who bears the consequences when those frameworks are too narrow,” he said.

Community narratives

The research found that while official discourse emphasised rapid response, “smart policing” and technological precision, community narratives focused more on fear, trauma, unemployment, spatial exclusion, and mistrust with policing bodies. He said rather than simply rejecting technology, many residents highlighted their alternative understanding of safety rooted in community agency, social investment and relational trust.

“Community members were not only questioning whether the technology works. They were also questioning whether violence should be framed primarily as a detection problem instead of a structural one,” Oosterwyk said.

In addition, he pointed out, imported surveillance technologies also carry embedded assumptions about risk, safety and governance that may not align with local realities. And the challenges posed by these systems are not primarily technical, but political.

“Dragnet surveillance systems do more than detect gunfire.”

“The key question is not only whether these systems function. But it is also whether they strengthen inclusive public reasoning and democratic legitimacy, or whether they narrow debate in ways that reproduce existing inequalities,” he said. “Dragnet surveillance systems do more than detect gunfire. They also shape who is perceived as dangerous, what counts as evidence, and where state resources are directed. Those decisions have redistributive consequences.”

Introduce stronger democratic safeguards

What’s needed, Oosterwyk said, is stronger democratic safeguards with procuring and deploying public-safety technologies. He recommended introducing independent performance audits, transparent reporting on false alerts and policing outcomes, participatory oversight mechanisms, community-defined safety indicators and greater investment in social infrastructure to address the root causes of violence.

Further, the paper also proposed a decolonial design review of imported policing technologies to evaluate the governance assumptions and accountability structures embedded in surveillance systems before deployment.

Ultimately, he said, ethical public safety governance requires more than better technology. He said it requires better democratic processes, stronger accountability institutions and meaningful community participation in decision-making that directly affect individuals’ lives.

The paper will be presented at the Pacific Asia Conference on Information Systems in Jakarta, Indonesia in July.