Attack Vectors

What and how generative AI techniques are used to deceive, impersonate and defraud.
Back to Attack Vectors

Deepfake robocalls

Audio
Consumer Fraud
Political Disinformation

Robocall deepfakes are automated mass calls that use cloned voices of public figures, executives, or generic personas to deliver pre-recorded messages at scale. The combination of synthetic voice and bulk dialing turns voice fraud into a one-to-many attack vector.

Use cases: Telco · Public sector

How it works

  • A target voice is cloned, or a generic synthetic voice is generated
  • Audio is paired with VoIP infrastructure that places thousands of calls per hour
  • Caller ID spoofing makes the calls appear to originate from local or trusted numbers

Where it shows up

  • Election interference using cloned candidate voices to suppress turnout or mislead voters
  • Political fundraising fraud impersonating campaign staff
  • Mass financial scams targeting consumers with cloned bank or agency voices
  • Disinformation campaigns issuing fake public safety alerts
  • Telecom fraud routing victims to bogus fraud-prevention hotlines