TY - JOUR
T1 - Three Levels of AI Transparency
AU - Haresamudram, Kashyap
AU - Larsson, Stefan
AU - Heintz, Fredrik
PY - 2023
Y1 - 2023
N2 - Transparency is generally cited as a key consideration towards building Trustworthy AI. However, the concept of transparency is fragmented in AI research, often limited to transparency of the algorithm alone. While considerable attempts have been made to expand the scope beyond the algorithm, there has yet to be a holistic approach that includes not only the AI system, but also the user, and society at large. We propose that AI transparency operates on three levels, (1) Algorithmic Transparency, (2) Interaction Transparency, and (3) Social Transparency, all of which need to be considered to build trust in AI. We expand upon these levels using current research directions, and identify research gaps resulting from the conceptual fragmentation of AI transparency highlighted within the context of the three levels.
AB - Transparency is generally cited as a key consideration towards building Trustworthy AI. However, the concept of transparency is fragmented in AI research, often limited to transparency of the algorithm alone. While considerable attempts have been made to expand the scope beyond the algorithm, there has yet to be a holistic approach that includes not only the AI system, but also the user, and society at large. We propose that AI transparency operates on three levels, (1) Algorithmic Transparency, (2) Interaction Transparency, and (3) Social Transparency, all of which need to be considered to build trust in AI. We expand upon these levels using current research directions, and identify research gaps resulting from the conceptual fragmentation of AI transparency highlighted within the context of the three levels.
KW - Artificial Intelligence
KW - Transparency
KW - Algorithm
KW - Interaction
KW - Society
KW - Governance
U2 - 10.1109/MC.2022.3213181
DO - 10.1109/MC.2022.3213181
M3 - Article
VL - 56
SP - 93
EP - 100
JO - Computer
JF - Computer
SN - 1558-0814
IS - 2
ER -