Computational Deception

In the future our daily life interactions with other people, with computers, robots and smart environments will be recorded and interpreted by computers or embedded intelligence in environments, furniture, robots, displays, and wearables. These sensors record our activities, our behaviour, and our interactions. Fusion of such information and reasoning about such information makes it possible, using computational models of human behaviour and activities, to provide context- and person-aware interpretations of human behaviour and activities, including determination of attitudes, moods, and emotions. Sensors include cameras, microphones, eye trackers, position and proximity sensors, tactile or smell sensors, et cetera. Sensors can be embedded in an environment, but they can also move around, for example, if they are part of a mobile social robot or if they are part of devices we carry around or are embedded in our clothes or body. Our daily life behaviour and daily life interactions are recorded and interpreted. How can we use such environments and how can such environments use us? Do we always want to cooperate with these environments; do these environments always want to cooperate with us? We argue that there are many reasons that human inhabitants of these environments do want to keep information about their intentions and their emotions hidden. Also their artificial interaction partner may have similar reasons to not give away all information they have or to treat their human partner as an opponent rather than someone that has to be supported. We survey situations where we can expect that human and artificial partner will not be honest to each other and will hide what they think, want or intend. These situations occur when the computer gets involved in social interaction, commerce and negotiation, and sports and games.