FARO (Frontline Aggregated Record of Online Harms) is a proof-of-concept tool built by SHIELD - APS, a nonprofit based in Italy, to address a gap in how online harms are handled. The people closest to the harm rarely hold enough consolidated evidence to be heard by the decision makers who could act on it. FARO uses a chain of AI agents to gather harm reports from many sources and turn them into clear trends, presented in one version for decision makers and one for the community that supplied the information. A person reviews each agent's work before the next agent begins, and the design puts anonymity, consent and community ownership of data first.
The case study covers how the pipeline was built, the data collection method, the practical lessons from building this tool and what comes next.
