Algorithms in general and Artificial Intelligence in particular, have the potential to radically transform individual lives and societies, “a transition comparable to (or more significant than) the agricultural or industrial revolution”. It is immensely uncertain, though, which of their capabilities will have the greatest impact. In any case, they pose both tremendous opportunities e.g. for improving health, wealth, sustainability, science, and self-understanding, as well as enormous risks, ranging from increasing economic inequality to discrimination, the loss of privacy, and the enabling of robust totalitarianism.

Their increasing pervasiveness in a growing number of domains has led to debates about and observations of a “governance by algorithms”, conceiving of algorithms as “one of the fundamental shaping devices of our daily, immersed-in-information existence”. As a response, the new field of “governance of algorithms” is emerging, aiming at reinforcing their benefits and minimizing their risks.

This may, however, prove difficult, e.g. due to particularities of the technology in question, its respective field of application, or limits of governance approaches and measures. As a result, their effectiveness or overall suitability may be impaired.

The Governance Topic Group reviews the existing literature on challenges and possible solutions, aiming at clustering issues that have been identified in various contexts, ranging from healthcare to democracy, with regards to different problems, such as freedom, fairness, or accountability.