Deepfake
Also called Synthetic media.
Audio, video or images generated or altered by AI to show someone saying or doing something they did not.
What it is
Media produced or manipulated with machine learning to convincingly imitate a real person's face or voice. NSA, FBI and CISA describe it as a growing tool for impersonation, fraud and disinformation against organizations.
Why it matters
It removes the old safety check of recognizing a voice or a face. Payment approvals, help desk resets and executive requests now need verification that does not depend on how someone sounds.
In practice
A finance employee joins a video call where the CFO and two colleagues ask for an urgent transfer. Every face on the call is synthetic.
What to look for
- Urgent requests for money or access delivered by voice or video only.
- Callers who avoid a call-back to a known number.
- Visual or audio glitches around the mouth, eyes or background.
Common mistake
Training people to spot artifacts. The artifacts are disappearing; a verification procedure through a separate, known channel is what holds.
Practice it nearby
CyberKit has no tool for this term itself. These are the closest places to practice it.
- Phishing Case
A reported invoice email: pull the deciding values out of its raw headers, and say what they establish.
- Email Header Analyzer
Walk the Received chain, and read the SPF, DKIM and DMARC verdicts.
Read next
- Vishing: Phishing carried out over a phone call instead of email or text.
- Business Email Compromise: Fraud that impersonates a trusted party in email to redirect money or data.
- Phishing: Tricking people into revealing credentials or running malware via fake messages.
Sources
- NSA, FBI and CISA: Contextualizing Deepfake Threats to Organizations
- CISA: NSA, FBI and CISA Release Cybersecurity Information Sheet on Deepfake Threats
- NIST AI 100-2 E2025: Adversarial Machine Learning, a Taxonomy and Terminology
- NIST AI Risk Management Framework
- MITRE ATLAS
- CISA, NSA and partners: Deploying AI Systems Securely