WP6- Social AI: Learning and Reasoning in Social Contexts


  • Sarit Kraus gave a talk entitled "Agent-Human Collaboration and Learning for Improving Human Satisfaction" on May 12th, 2021.


The main goal of WP6 is to promote the study of foundations, techniques, algorithms and tools for allowing autonomous AI agents to be social, that is, to cooperate, interact, negotiate, learn, and thus act in societies. The major challenge here is considering that AI agents will need to act in social contexts and their actions will influence the actions of others in such societies. This, therefore, includes learning from others, interacting, cooperating and negotiating in (possibly massive) multiagent systems and hybrid systems with populations that may have both agents and humans.

We envision a world where typical devices will become more intelligent (agents), mobile (robots) and able to communicate, and thus socialize. This is a multidisciplinary endeavour, involving areas such as agent communication, multiagent planning, multiagent reinforcement learning, automated negotiation, complex systems and social simulations.

This work package will establish a roadmap in this field drawing new directions for research exploring new avenues for social AI focusing on its foundation's techniques and algorithms.


The key questions that will drive the research on the fundamentals of social AI are:

  • How do we empower individual AI agents to communicate with each other, collaborate, negotiate and reach agreements?

  • How can agents coordinate to fairly share common resources?

  • How can we make agents learn from each other in a responsible and fair way, leading to more intelligent behavior?

  • How to create trustworthy hybrid human-AI societies that fulfil humans’ expectations and follow their requirements?


The work in WP6 is divided into four ‘scientific challenge tasks’, i.e., addressing the main scientific challenges in the theme, plus two extra tasks, one on cross-fertilization with industry and one on fostering a scientific community dedicated to this theme. The scientific challenges listed previously will be addressed in these four tasks where the community will work to shape the research worldwide. Each scientific task will continuously interact with the other tasks in order to provide input and receive feedback and challenges and thus get a closed-loop approach to address the large challenge that underpins Social AI. Notice that the scientific challenges will have strong synergies with WP2, WP3, WP4 and WP5.

Task 6.1

Modelling social cognition, collaboration and teamwork

Task 6.2

Theoretical models for cooperation between agents

Task 6.3

Learning from others

Task 6.4

Emergent Behaviour, agent societies and social networks

Task 6.5

Synergies Industry, Challenges, Roadmap on social AI system

Task 6.6

Fostering the AI scientific community on the theme of social AI