Verification Procedures and Callbacks
What you'll learn: How to protect your organization by independently verifying identities before granting sensitive requests—the last line of defense against social engineering.
The Core Problem
Even with excellent security awareness training and phishing simulations (which you've already learned), attackers can still succeed by impersonating trusted individuals or using stolen credentials. The solution? Never trust the communication channel the request came through—always verify through a separate, independent channel you control.
How Verification Callbacks Work
Think of it like your bank: when you call claiming you forgot your password, they don't just reset it on that same call. They hang up and call you back at the number on file, or send verification to a pre-registered email. You're proving identity through a channel the attacker can't control.
Essential Callback Protocol Elements
1. Independent Channel Selection
- Don't use phone numbers or contact details provided in the suspicious request
- Use only verified contacts from your internal directory or records
- Never call back the number on caller ID—look it up yourself
2. Time-Based Verification
- Implement mandatory delays for sensitive changes (password resets, wire transfers)
- This cooling-off period disrupts the attacker's urgency tactics you learned earlier
3. Multi-Person Authorization
- Require two separate verifications for high-risk actions
- Example: IT verifies with the user; finance verifies with IT
4. Documentation Requirements
- Log all verification attempts and callbacks
- Integrate with your physical access logging systems for complete audit trails
Real-World Application
When someone calls IT requesting a password reset: don't reset it immediately. Hang up, call the employee back at their registered extension, confirm the request, then proceed. This simple step defeats most vishing attacks.
Key Takeaway: Verification callbacks transform trust-based systems into zero-trust identity verification by requiring independent confirmation through channels you control, not the attacker.