Minority Report and AI: are we moving towards predictive policing?

When science fiction starts to feel familiar

In Minority Report, authorities act on predictions of crimes before they happen. Dr Joseph Opare uses the film as the starting point for an unsettling question: could modern surveillance and predictive policing take society down a similar path?

The parallel is about what happens when a prediction begins to shape how someone is treated. If a system labels a person a potential threat, how much opportunity do they have to question that judgement before it affects their life?

The film is fiction, but the episode argues that developments in technology make its central dilemma increasingly relevant. The concern is the direction of travel: moving from investigating what someone has done towards judging what a system thinks they might do.

Palantir and the predictive-policing debate

From that film comparison, the episode turns to Palantir’s technology and the wider debate about AI-driven surveillance and policing. It asks whether systems that analyse behaviour and patterns could improve safety, and what their use might mean for privacy and freedom.

It also asks whether similar systems could be adopted in Africa. That places the debate within decisions about public oversight and the terms on which governments introduce new technology.

A prediction is a use of past information

Predictive policing uses data to estimate where an offence might occur or who might require attention. It does not observe a future event. Its results depend on the information selected, the assumptions built into the system and how officials use its output.

That makes the source of the data important. A record of police activity is not necessarily a complete record of crime. Some incidents go unreported, while some places receive more scrutiny. Feeding those differences into software can reproduce them in a new form.

Technology does not remove responsibility

UNESCO’s recommendation on AI ethics addresses privacy, discrimination and oversight. Those concerns become especially concrete when a system influences how public authorities treat a person.

A useful safeguard starts with a plain explanation of what a tool does. Does it suggest patrol locations, flag records for review or influence a decision about an individual? These are different uses with different consequences. Calling all of them “AI” can hide those distinctions.

For Black communities, scrutiny is not abstract

The episode asks what these systems could mean for Africa and the diaspora. The central concern is whether an existing pattern of unequal attention becomes harder to question once it carries a technical label.

Our analysis is that communities need ways to challenge both the data and the decision. Officials should be able to explain why a tool is useful, what errors it makes and what happens when its output is disputed. Human involvement only helps if the person reviewing a result can meaningfully reject it.

Ask what success means

More alerts do not automatically mean greater safety. A public discussion should examine whether the system helps solve real problems, how often it points in the wrong direction and whether its benefits justify the intrusion.

That brings the discussion back to Minority Report: how much power should a prediction have over a person’s freedom? The episode asks whether society is examining that question quickly enough as the technology develops.