An independent research group that maintains a public, sourced register of documented AI-enabled fraud, deepfake impersonation and agentic exploitation. Nine ranked risk categories, 44 cited sources. Every figure is linked to the government publication, peer-reviewed paper, industry disclosure or incident archive that produced it. No synthetic or composite cases. Registry 2026.1, revised 27 October 2026.
Cite this register: AI Safety Research Group, "Synthesized-voice family extortion," Risk registry 2026.1, updated 27 October 2026, https://aisafetyresearchgroup.com/risks/synthesized-voice-family-extortion.