V.O.I.C.E.: A Data-Driven Risk Taxonomy for Synthetic Voice Generation
summary
V.O.I.C.E. is a data-driven taxonomy built from 569 incidents, 2,221 Reddit discussions and 1,069 direct reports that classifies six risk categories and maps concrete attack chains: voice harvesting from public talks, $300 services producing 1-minute clones, pairing synthesized audio with stolen PII and face data to pass remote interviews, and exploitation that escalates into child abuse material or targeted harassment. It shows contextual drivers—degree of exposure, social visibility, legal resources—and evaluates mitigation gaps like disclosure that does not equal agency and tools such as Origin Story hardware-rooted provenance for detection of remote hiring fraud.
AI-written summary · about
chapters (5)
- 0:00Introducing V.O.I.C.E.
- 2:59AI voice regulations
- 8:17Building voice risk taxonomy
- 15:03Next steps for voice risk
- 17:32Q&A
related talks
39mSynthetic Trust: Exploiting Biases at Scale
36mFabricated Realities: AI Personas and Simulacra and the Threats They Pose
35mBuilding Trust & Enabling Innovation for Voice Enabled IoT
16mFrom Existential to Existing Risks of Generative AI: A Taxonomy of Who Is at Risk,...
45mOverview of Risks, Vulnerabilities and Paths to Safer Use of Generative AI
36mDemocratized Deception: How Threat Actors Are Weaponizing AI to Scale Cybercrime
39m Put Words In My Mouth
58mSTOIC Security: Shielding Your Generative AI App from the Five Deadly Risks