MSA: A Persuasion-Grounded Taxonomy of Elder-Fraud Mechanisms, Signals, and Archetypes

Elder-fraud defense has no shared vocabulary. Regulators count scam categories. Law enforcement counts crime types. Industry groups count phishing campaigns. Gerontology explains how persuasion works. These descriptions have been connected before, most systematically by Beals and colleagues, but none is designed to drive detection while an encounter is still running. Cyber defense had the same problem until MITRE ATT&CK gave defenders a common language for attacker behavior. This paper introduces MSA (Mechanism, Signal and Archetype), a three-level taxonomy of elder fraud. A Mechanism is the persuasion lever a scam pulls, taken from influence and victimization theory. A Signal is one observable behavior, such as a demand for gift cards. An Archetype is a named scam family: a combination of signals, mapped to the categories that the Federal Trade Commission (FTC), the FBI's Internet Crime Complaint Center (IC3), and the Anti-Phishing Working Group (APWG) already count. Two properties set MSA apart from earlier fraud classifications. First, every signal states how it can be observed: from what the victim says or from observing and analyzing artifacts involved in the scam. Second, MSA is a maintained artifact, not a table in a paper. It ships as versioned, machine-readable releases with stable identifiers, kept current by a pipeline in which language models propose updates and a human expert decides. We present the design requirements, the structure, the current library, six use cases, and the maintenance model. The library described here, release 0.2.0, is published separately under CC BY 4.0 and archived at https://doi.org/10.5281/zenodo.22085666.