Panel: The future of trust modeling research

There has been a TRUST workshop at AAMAS for over 20 years now.
The multiagent systems community has featured papers on trust
at its conferences over the years as well, with a varied focus
of attention, from issues of representing reputation through
an aggregation of opinions, incentivizing honesty, considering
attacks and collusion, designing testbeds for validation,
studying logical and philosophical foundations, coping with
a lack of information, and imagining a host of valued applications.
Of late, the entire AI community has been challenged to consider,
from the very beginning, the issue of whether the systems which they
design will be trusted by their human users. This has led to a variety
of specific areas of concern, each with a host of approaches and
solutions (including ones examining fairness, accountablity and
transparency and others mapping out important consideration for
human-agent partnerships). These themes have become showcase items
at recent international conferences.

Looking towards the future, two important questions arise:
what is the future of trust modeling research (the most challenging
and interesting concerns to address) and how can our community
initiate a proper dialogue with other AI researchers who are now
attuned to issues of trust. For the latter topic, two companion 
questions arise: Can trust modeling paradigms assist
in ensuring trusted AI? Do trusted AI solutions provide important
insights for trust modeling?

This panel will look towards the future of trust modeling research.
Each panelist will be asked to expound on what they perceive to
be the most valued steps forward with research, grounding their
discussion in part in their own research AND to comment on their
conclusions about the relationship between our community and that
of what can be viewed as trusted AI.

Each panelist will have 10 minutes to speak. This will be followed
by an opportunity for cross-talk and questions from the audience.
All workshop participants are urged to come with specific questions
for the panelists, even before the position statements have been made.
We are all responsible for mapping out an exciting future for our field.

Panelists:

Sandip Sen (University of Tulsa)
Munindar Singh (North Carolina State University)
Katia Sycara (Carnegie Mellon University)
Moritz Hardt (University of California, Berkeley)
Comments