Algorithmic Transparency and Manipulation

Michael Klenk

DOI: https://doi.org/10.1007/s13347-023-00678-9

2023-11-22

Abstract:A series of recent papers raises worries about the manipulative potential of algorithmic transparency. But while the concern is apt and relevant, it is based on a fraught understanding of manipulation. Therefore, this paper draws attention to the indifference view of manipulation, which explains better than the vulnerability view why algorithmic transparency has manipulative potential. The paper also raises pertinent research questions for future studies of manipulation in the context of algorithmic transparency.

Artificial Intelligence

What problem does this paper attempt to address?

This paper attempts to explore the relationship between algorithmic transparency and manipulation, and suggests that current discussions on algorithmic transparency are insufficient. Specifically, the paper focuses on the following points:

1. **The Potential Manipulativeness of Algorithmic Transparency**:
- The paper points out that although algorithmic transparency is generally seen as a key element in the responsible design and deployment of AI systems, this transparency may have underrecognized negative impacts, namely that it can be used to manipulate users.

2. **Issues with the Vulnerability Perspective**:
- The current understanding of the relationship between algorithmic transparency and manipulation is mainly based on the "vulnerability perspective," which posits that manipulation exploits individuals' vulnerabilities. However, this perspective itself is problematic because it fails to clearly explain how and why algorithmic transparency constitutes manipulation.

3. **Introduction of the Indifference Perspective**:
- The paper introduces another perspective—the "indifference perspective"—to better explain why algorithmic transparency has the potential for manipulation. According to this perspective, when transparency is applied indiscriminately without providing reasons to users, it can easily evolve into manipulative behavior.

4. **Future Research Directions**:
- The paper also proposes some future research directions, emphasizing the need to further explore the potential for manipulation due to algorithmic transparency in different contexts, and calls for scholars to conduct more in-depth research on this issue.

In summary, this paper aims to address the shortcomings in the existing theoretical framework by introducing the "indifference perspective" and provides new insights for further research on the complex relationship between algorithmic transparency and manipulation.