An Ethical eValuation Agent (EeVA): Results of a Proof-of-Concept Test on a Prototype Agentic-like Workflow to Assist Ethical Deliberations

The article presents EeVA, an LLM-based workflow designed to assist in ethical deliberations by evaluating use cases against ten ethical frameworks. Built using n8n, EeVA operates through three workflows—starter, worker, and emitter—producing structured evaluations and syntheses that highlight ethical tensions and recommend modifications without providing definitive answers. This approach aims to support non-expert personnel in navigating complex ethical challenges, emphasizing the need for further research on reproducibility and user testing to enhance its utility in practical applications.