Shoshana Zuboff’s concept of surveillance capitalism, developed in her 2019 book of the same name, represents one of the most important theoretical frameworks for understanding the political economy of contemporary AI. Though Zuboff does not identify as a Marxist, her analysis shares important structural features with Marxist political economy, and has been extended and critiqued by Marxist scholars in productive ways.

Zuboff argues that surveillance capitalism is a new economic logic that emerged from Google’s discovery, in the early 2000s, that the data generated by user behaviour on its search platform could be used not just to improve the platform, but to predict and influence user behaviour for advertising purposes. This discovery, that human experience could be rendered as data, processed by AI, and converted into prediction products sold to advertisers, constitutes a new means of capital accumulation.

Behavioural data as raw material. In surveillance capitalism, human experience is the raw material from which value is extracted. Every search, click, like, location check-in, and purchase generates behavioural data that platforms capture and process. This data is not used primarily to serve the user, it is used to build models of user behaviour that allow advertisers to target their messages with unprecedented precision.

Prediction products. The output of surveillance capitalism’s production process is not a service. It is a prediction, a probabilistic statement about what a given user will do, buy, or believe. These predictions are sold to advertisers, political campaigns, insurance companies, and anyone else willing to pay for the ability to influence human behaviour at scale.

The modification of behaviour. Zuboff’s most disturbing insight is that surveillance capitalism does not merely predict behaviour, it actively seeks to modify it. The AI systems that underlie social media platforms, recommendation engines, and targeted advertising are not neutral conveyors of information. They are behaviour modification systems, designed to maximise engagement, attention, and ultimately consumption.

Marxist critique and extension. Marxist scholars have extended Zuboff’s framework in important ways. Nick Dyer-Witheford and others argue that surveillance capitalism is not a departure from capitalism but its latest development, and that the concept of digital labour (users as unpaid workers whose activity generates value for platforms) connects it more firmly to classical Marxist analysis than Zuboff’s framework suggests. The user who generates data is not merely having their experience commodified, they are labouring for the platform without compensation.

Reflection question: How does your digital behaviour generate value for platforms you use? Who profits from that value? What would fair compensation look like?