The field of algorithmic decision-making is shifting towards a more nuanced understanding of the complex interactions between multiple stakeholders. Researchers are developing innovative frameworks to address the limitations of individual-centric approaches, instead opting for system-level designs that prioritize social welfare and collective feasibility. A key direction in this area is the integration of human oversight and AI-driven decision-making, enabling more effective and inclusive outcomes. Notable papers in this area include:

• A production-ready machine learning system for inclusive employment, which demonstrates the potential for AI-driven disability job matching platforms to increase service capacity while maintaining human oversight.
• FairVizARD, a visualization system for assessing multi-party fairness in ride-sharing matching algorithms, which provides a valuable tool for evaluating and balancing the fairness of different parties involved.
• Collaborative matching, a data-driven algorithmic matching system that takes a collaborative approach to human-AI decision-making, has shown promising results in achieving human-AI complementarity and improving matching outcomes.