Workerville: Towards an Organizational Behavior Account of Agent Safety

Key points

• LLM-based agents now interact with their environments continuously, shaped by such organizational channels as user instructions, peer messages, and long-term memory.1
• How such factors jointly shape an agent's safety behavior from a unified perspective remains unmeasured.1
• To bridge this gap, we advocate organizational behavior (OB) as a framework for studying the safety of advanced agents, reorganizing the objects of study, theoretical foundations, and experimental design around the relational structure in which agents operate.1
• To operationalize ACB, we introduce Workerville, a controlled benchmark that manipulates organizational conditions over shared tasks, applying 16 organizational configurations to 210 tasks to yield 3,360 challenges, evaluated by human-validated agentic judges.1

Sources (1)

• [1]Workerville: Towards an Organizational Behavior Account of Agent Safety
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 09:23 AM

  We present the first systematic formalization of counterproductive work behavior (CWB), a canonical safety-relevant subfield of OB, as Agentic Counterproductive Behavior (ACB).

  LLM-based agents now interact with their environments continuously, shaped by such organizational channels as user instructions, peer messages, and long-term memory.

Extractive summary: sentences quoted from the sources.