In 1791, philosopher Jeremy Bentham designed the Panopticon: a circular prison where a single guard could observe all inmates without them knowing whether they were being watched at any given moment. The idea was that the mere possibility of being observed would induce self-discipline. Two centuries later, Michel Foucault used the Panopticon as a metaphor for modern disciplinary societies. Today, with artificial intelligence, we have gone further: we don't just watch, we can analyse, predict and classify the behaviour of entire populations in real time.
Traditional surveillance had an inherent limitation: cost. Each person under surveillance required resources — agents, equipment, time. The Stasi, East Germany's secret service, employed one informant for every 6.5 citizens and still could not monitor everyone. AI has removed this constraint. A single camera system with facial recognition can identify thousands of people per minute. A data traffic analysis algorithm can monitor the communications of millions of users simultaneously. The marginal cost of surveillance has collapsed towards zero.
The most extreme case is the Chinese social credit system, which combines facial recognition, financial data analysis, social media monitoring and GPS tracking to assign a "score" to citizens. A low score — caused by late payments, jaywalking, associating with people who have low scores — can prevent buying plane tickets, enrolling children in certain schools, accessing credit. It is Bentham's Panopticon realised on a national scale.
But it would be a mistake to think mass surveillance is only a Chinese problem. Edward Snowden's 2013 revelations demonstrated that the American NSA was massively collecting phone metadata and intercepting internet communications globally. In Europe, predictive policing (PredPol, HunchLab) uses historical crime data to "predict" where future crimes will occur — with the predictable result of concentrating surveillance in poor neighbourhoods and communities of colour. In Italy, the SARI system (Automatic Image Recognition System) was used by law enforcement without a clear legal basis.
Mass surveillance is not just a privacy issue. It is a question of power. Who can surveil whom? Who controls the collected data? Who decides how it is used? When surveillance is asymmetric — governments and corporations watch us, but we cannot watch them — the power relationship is fundamentally unbalanced. Privacy is not the right to "have nothing to hide". It is the right to exist without being constantly judged, classified and manipulated.