Voice Security - Get Up to Speed on This Attack Vector
summary
Explains that digital signing via STIR/SHAKEN reduces spoofing but fails when calls traverse PSTN because signatures drop and it needs end‑to‑end IP; describes analytics and reputation monitoring that score numbers (crowdsourced labels can misclassify legitimate contact center calls); outlines enterprise voice firewalls and layer‑7 TDoS detection, contact‑center audio forking with device acoustic fingerprints and PIN‑timing checks, and voice biometrics plus deepfake detection (Pindrop flags a Biden deepfake in seconds); recommends outbound controls: anti‑spoofing APIs and call branding with signed rich call data.
AI-written summary · about
chapters (8)
- 0:00Voice security threats
- 7:24Regulatory efforts against fraud
- 15:07Voice security vs assurance tools
- 21:30Contact center fraud tactics
- 26:26Outbound voice assurance solutions
- 32:06Start with voice security
- 35:15Deepfake detection future
- 40:13Q&A
related talks
49mChipmunk or Pepe? Using Acoustical Analysis to Detect Voice-Channel Fraud
26mThe Forgotten
39mSynthetic Trust: Exploiting Biases at Scale
36mFabricated Realities: AI Personas and Simulacra and the Threats They Pose
33mYour voice confirms my identity
20mSecurity in the Age of AI: Response and Strategy
39m Put Words In My Mouth
29mAuthentiCall: Efficient Identity and Content Authentication for Phone Calls