Examples of themes and projects that we are interested in addressing include (but are not limited to):

• Language models are increasingly used to draft court rulings, shifting judges’ roles from writing to supervising. We aim at developing a benchmark to test whether LLMs can produce legal reasoning comparable to trained lawyers. (PIs: Engel, Gummadi)
• In the courtroom, LLMs are most powerful if the entire judicial workflow is digitized. We investigate in which ways electronic files make rulings vulnerable to purposeful manipulation, and how courts can protect their integrity. (PIs: Cha, Engel)
• This initiative develops synthetic versions of users’ ‘future selves’ as an intervention point. Using virtual reality technology in combination with AI-powered LLMs, realistic avatars of users are generated to enable meaningful interaction between them and their future self to improve their ability for future-oriented thinking (PIs: Cha, Gummadi, Van Gelder, Zagheni)
• Virtual replicas of public spaces such as train stations, squares, and neighborhoods—“virtual twins”—will be developed. AI will support the creation and modification of these environments, enabling realistic variations in features such as lighting, crowd density, and signs of disorder. This approach enables rigorous theory testing and helps identify how public safety interventions can reduce anti-social behavior and improve safety (PIs: Cha, Gummadi, Van Gelder)
• How have Artificial Intelligence and digital technologies changed the way people access information and services, communicate and interact? And what are the implications for gender and social inequalities, and for governing society?  To address these questions, the project would also develop methods to combine data donation within solid statistical frameworks to generate representative observatories of online behavior. (PIs: Gummadi, Cha, Zagheni, Engel)
• Rapid advances in Machine Learning and Artificial Intelligence are opening up new opportunities to model and predict socio-demographic and economic indicators over the life course. This project would combine powerful predictive tools with approaches from causal inference to advance our understanding of socio-economic inequalities. (PIs: Gummadi, Cha, Zagheni)
• As agentic AI integrates human workers into automated workflows, it becomes important to understand the AI-human matching process under demographic bias and skillsets. This project will examine how agentic AI reshapes task allocation and labor valuation, providing critical evidence for policymakers navigating the AI-driven economy. (PIs: Cha, Gummadi)

Role and Research Environment

As a postdoc in this network, you will:

• Conduct Independent Research: Lead research projects that bridge social science inquiry with the engineering of digital artefacts.
• Collaborate Across Institutes: While each postdoc will have a primary location at one of the participating institutes, you are expected to spend substantial time visiting and collaborating with other members across the network to facilitate seamless exchange and cross-fertilization of ideas.
• Participate in Network Activities: Engage in a vibrant network with a central hub at the MPI for Software Systems in Kaiserslautern to foster cross-pollination of ideas and shared technical solutions.

Program Benefits & Terms

The positions are part of the Max Planck Postdoc Program, offering world-class support for early-career researchers:

• Contract & Salary: Fixed-term, 3-year employment contracts based on the German public service wage agreement (TVöD Bund).
• Professional Development: Access to the Planck Academy for targeted mentoring, career coaching, and seminars on leadership, project management, and science communication.
• Research Funding: A generous travel allowance  for conferences, workshops, and international networking.
• Work-Life Balance: Support for mobile work (subject to supervisor agreement and institute’s guidelines), access to childcare facilities facilitated by the institutes, and comprehensive social security benefits (health, pension, and unemployment insurance).