Does Your Awareness Training Platform Work Against AI Powered Social Engineering Attacks?

Do most awareness training platforms protect against deepfakes?

No. Most awareness training platforms do not prepare employees for AI-powered social engineering attacks.

What does research show about human deepfake detection?

Key finding: Training employees to visually spot deepfakes does not work. Research shows human detection accuracy is not significantly above chance.

A 2024 meta-analysis by Diel et al., synthesising 56 peer-reviewed studies involving 86,155 participants, found:

Source: Diel et al. (2024), Human performance in detecting deepfakes: A systematic review and meta-analysis, Computers in Human Behavior Reports.

Critical implication: Effective training must focus on verification processes and behavioural responses, not visual identification.

What does Breacher.ai assessment data indicate?

Note: The following is proprietary internal assessment data, not independently verified peer-reviewed research.

• Organisations with mature awareness training show measurably lower attack success rates.
• Multi-channel staged attacks consistently bypass single-channel defences.
• Finance and HR departments show elevated vulnerability due to transaction authority.

How should you evaluate your awareness training platform?

• Does it simulate voice-based attacks, or only email?
• Can it simulate video conference deepfakes on Teams/Zoom?
• Does training cover verification processes for urgent requests?
• Does training teach visual deepfake detection? (Red flag – research shows this does not work)
• Does it measure action rates (objectives achieved), not just click rates?

Sources

• Diel et al. (2024). Human performance in detecting deepfakes: A systematic review and meta-analysis. Computers in Human Behavior Reports. DOI: 10.1016/j.chbr.2024.100471
• iProov (2025). Deepfake Blindspot Study.
• Proofpoint (2024). State of the Phish Report.
• Breacher.ai internal assessment data (proprietary).