Vassilis-Javed Khan

Maurits Kaptein, Vassilis-Javed Khan, Andriy Podstavnychy

AI agents -- systems that plan, reason, and act using large language models -- produce non-deterministic, path-dependent behavior that cannot be fully governed at design time, where with governed we mean striking the right balance between as high as possible successful task completion rate and the legal, data-breach, reputational and other costs associated with running agents. We argue that the execution path is the central object for effective runtime governance and formalize compliance policies as deterministic functions mapping agent identity, partial path, proposed next action, and organizational state to a policy violation probability. We show that prompt-level instructions (and "system prompts"), and static access control are special cases of this framework: the former shape the distribution over paths without actually evaluating them; the latter evaluates deterministic policies that ignore the path (i.e., these can only account for a specific subset of all possible paths). In our view, runtime evaluation is the general case, and it is necessary for any path-dependent policy. We develop the formal framework for analyzing AI agent governance, present concrete policy examples (inspired by the AI act), discuss a reference implementation, and identify open problems including risk calibration and the limits of enforced compliance.

Ioanna Lykourentzou, Federica Lucia Vinella, Faez Ahmed et al.

As the volume and complexity of distributed online work increases, the collaboration among people who have never worked together in the past is becoming increasingly necessary. Recent research has proposed algorithms to maximize the performance of such teams by grouping workers according to a set of predefined decision criteria. This approach micro-manages workers, who have no say in the team formation process. Depriving users of control over who they will work with stifles creativity, causes psychological discomfort and results in less-than-optimal collaboration results. In this work, we propose an alternative model, called Self-Organizing Teams (SOTs), which relies on the crowd of online workers itself to organize into effective teams. Supported but not guided by an algorithm, SOTs are a new human-centered computational structure, which enables participants to control, correct and guide the output of their collaboration as a collective. Experimental results, comparing SOTs to two benchmarks that do not offer user agency over the collaboration, reveal that participants in the SOTs condition produce results of higher quality and report higher teamwork satisfaction. We also find that, similarly to machine learning-based self-organization, human SOTs exhibit emergent collective properties, including the presence of an objective function and the tendency to form more distinct clusters of compatible teammates.