Many of the challenges we face, such as loss of biodiversity or air pollution, require collective action. But people often fail to work together for the common good due to problems of co‑ordination. The costs to individuals of participating in collective action include, for example, having to give up power or change their lifestyle or the risk that others will benefit unfairly as a result of their efforts (known as free-riding).

The Artificial Intelligence Lab at the Vrije Universiteit Brussel is carrying out experiments to determine how the presence of artificial agents affect group decision-making. Artificial agents are complex pieces of software that have the ability to act autonomously based on a pre-programmed set of ‘beliefs’. By interacting directly with human participants, they can influence and change activity to achieve specific goals. The experiments of the Artificial Intelligence Lab are testing whether delegating tasks that require acting for the public good to artificial agents improves the ability of a group to navigate collective risk dilemmas.

This experiment is an example of how humans and artificial agents work together to solve a problem. The results will shed light on whether artificial agents can help improve co‑ordination and fairness among group participants in collective risk scenarios and therefore increase the group’s collective intelligence to solve complex problems. It also explores the important issue of trust: how willing are humans to delegate a decision to artificial agents, and will the collective be better off?