Title
Are we being manipulated by chatbots?
NB: This talk will be held in German
Abstract
In my talk, I argue that we should refrain from describing chatbots as if they were capable of manipulating us. Strictly speaking, only persons can manipulate other persons. A chatbot can help someone achieve a manipulative goal, but it is only in this derivative sense that chatbots can be said to manipulate. The proposed restriction on how we talk about chatbots is important for two reasons. First, when we use the term “manipulation,” we typically have ethically problematic behavior in mind. But chatbots cannot be held morally responsible. Responsibility lies with the people who design and supply them, and perhaps also with those in positions to regulate them but who fail to do so. Second, it is sensible to use concepts that are as clear as possible when describing AI and the problems it might create. Too often, however, the concept of manipulation is used even though more appropriate descriptions are available to us that would contribute to a better understanding of the situation. Talk of manipulation may then fail to illuminate the problem at hand. Worse still, it may inflate the problem, or even obscure facts about responsibility.
About Christiane
Dr. Christiane Turza is a Research Associate at the Chair of Practical Philosophy under Prof. Dr. Weyma Lübbe at the University of Regensburg. Her research focuses on the ethics of communication. Her doctoral dissertation was published in 2023 by Mentis under the title “Manipulation. Zum Begriff und ethischen Status.”