TY - GEN
T1 - Ethical Challenges in Data-Driven Dialogue Systems
AU - Henderson, Peter
AU - Sinha, Koustuv
AU - Angelard-Gontier, Nicolas
AU - Ke, Nan Rosemary
AU - Fried, Genevieve
AU - Lowe, Ryan
AU - Pineau, Joelle
N1 - Publisher Copyright:
© 2018 ACM.
PY - 2018/12/27
Y1 - 2018/12/27
N2 - The use of dialogue systems as a medium for human-machine interaction is an increasingly prevalent paradigm. A growing number of dialogue systems use conversation strategies that are learned from large datasets. There are well documented instances where interactions with these system have resulted in biased or even offensive conversations due to the data-driven training process. Here, we highlight potential ethical issues that arise in dialogue systems research, including: implicit biases in data-driven systems, the rise of adversarial examples, potential sources of privacy violations, safety concerns, special considerations for reinforcement learning systems, and reproducibility concerns. We also suggest areas stemming from these issues that deserve further investigation. Through this initial survey, we hope to spur research leading to robust, safe, and ethically sound dialogue systems.
AB - The use of dialogue systems as a medium for human-machine interaction is an increasingly prevalent paradigm. A growing number of dialogue systems use conversation strategies that are learned from large datasets. There are well documented instances where interactions with these system have resulted in biased or even offensive conversations due to the data-driven training process. Here, we highlight potential ethical issues that arise in dialogue systems research, including: implicit biases in data-driven systems, the rise of adversarial examples, potential sources of privacy violations, safety concerns, special considerations for reinforcement learning systems, and reproducibility concerns. We also suggest areas stemming from these issues that deserve further investigation. Through this initial survey, we hope to spur research leading to robust, safe, and ethically sound dialogue systems.
KW - Adversarial Examples
KW - Bias
KW - Computers and Society
KW - Dialogue Systems
KW - Ethics and Safety
KW - Machine Learning
KW - Natural Language Processing
KW - Privacy
KW - Reinforcement Learning
KW - Reproducibility
KW - Security
UR - http://www.scopus.com/inward/record.url?scp=85061026082&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85061026082&partnerID=8YFLogxK
U2 - 10.1145/3278721.3278777
DO - 10.1145/3278721.3278777
M3 - Conference contribution
AN - SCOPUS:85061026082
T3 - AIES 2018 - Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and Society
SP - 123
EP - 129
BT - AIES 2018 - Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and Society
PB - Association for Computing Machinery, Inc
T2 - 1st AAAI/ACM Conference on AI, Ethics, and Society, AIES 2018
Y2 - 2 February 2018 through 3 February 2018
ER -