Signal Assessing Company Contributions to Societal Resilience: Extending the Societal Capacity Assessment Framework to Agentic AI

Summary

Companies that build and deploy AI agents act as institutional players shaping the sociotechnical conditions under which this technology becomes embedded in existing social and economic systems. Because of this influence, they are positioned to help determine society's capacity to absorb and govern the consequences of agentic AI. The research argues that by designing and promoting safer agent implementations, these companies can actively build societal resilience rather than merely benefit from it. To put this idea into practice, the study reworks an existing indicator-based societal capacity assessment framework, inverting its original purpose of measuring resilience as a backdrop for deployment decisions into a tool for measuring how deployment decisions themselves affect resilience. The approach unfolds in two stages, first translating the framework's vulnerability, coping and adaptive capacity concepts into indicators tailored to agentic AI deployment, then applying that indicator set in a structured review of a major technology company's publicly available documentation.