AI-Driven Contextual Communication: Implications for Citizens and Society
- Contact person:
- Sara Leckner
- Marianne and Marcus Wallenberg Foundation
- Responsible at MaU:
- Sara Leckner
- Collaborators and other project members:
- Johanna Björklund - Umeå University
- Jesper Strömbäck - Gothenburg University
- Richard Wahlund - Stockholm School of Economics
- Time frame:
- 01 January 2021 - 31 December 2023
- Research subject:
About the project
This project studies the consequences for citizens and society when artificial intelligence becomes the dominating method of placing messages in online media contexts. We take a consumer perspective and study user behaviour to understand the implications for individuals, and by extension, for society as a whole.
Our research questions concern
- matters of privacy, integrity, and trust;
- communication efficiency in terms of information transfer, conversion rates, brand recall, etc.;
- the individuals' self-expected and actual response to contextual communication; and
- the functioning and soundness of democratic processes.
In each case, we contrast contextual communication with personalised communication. Simply put, the former matches messages with media contexts, whereas the later matches messages with recipients. Personalized advertising is currently the dominating form in digital advertising, but has severe problems related to personal integrity, stereotyping, and democracy. Contextual communication is expected to solve many of these problems, but may in doing so introduce new and potentially worse problems. The project establishes contextual communication as a new research field. It is a multidisciplinary undertaking, engaging researchers from computer science, behavioural science, social science, and economy. We use a combination of quantitative and qualitative methods, conducting experiments and longitudinal data collection on real-world advertising platforms, and complement these with studies involving, e.g., eye-tracking, walk-through techniques, and emotional response rates. The scientific outcome is an empirically validated theory of AI-driven contextual communication.