CYBERDIVA LABS
AI RISK LEARNING UNIT #001

VOICE
CLONING

Learn to recognize, assess and navigate an AI-era risk through one interactive case.

01 · EXPERIENCE
02 · DECONSTRUCT
03 · ASSESS
04 · PROTECT
05 · BUILD YOUR PLAN
06 · PRACTICE
01 · EXPERIENCE

MAYA ❤️

“Mom? I was in an accident. Don't call Dad. I need $2,000 right now.”
02 · DECONSTRUCT

What made the attack believable?

🎤 SOURCE AUDIOPublic audio can provide material for imitation.
🧬 AI CAPABILITYA voice model can generate new speech with similar vocal characteristics.
🌐 EXPOSUREPublic details provide believable context.
⏱️ HUMAN PRESSUREFear and urgency reduce time to verify.
AI enabled the imitation. It did not create the whole attack.
03 · ASSESS

Where was the risk?

Select every element that contributed.

04 · PROTECT

Build your defense

Select the controls you think are strong. The answer is not shown until you test your choices.

05 · BUILD YOUR PLAN

Create a family verification plan

Think of your phrase offline. Do not type your real family secret into CyberDiva.

1 · CHOOSE A PRIVATE PHRASENot a pet name, birthday, school, or something posted online.
2 · AGREE WHEN TO USE ITUnexpected money requests, emergencies, new numbers, or unusual sensitive requests.
3 · VERIFY INDEPENDENTLY ANYWAYEnd the suspicious contact and use a number or channel you already trust.
4 · REPLACE IT IF EXPOSEDTreat the phrase like a shared secret.
Defense in depth: a family phrase is an extra layer — not a replacement for independent verification.
06 · PRACTICE

Try the call again

“Mom, please. Don't call anyone. I need the money now.”

Choose the safest actions in order.

Build the sequence.
DIGITAL TRUST SKILL UNLOCKED

AI Risk: Independent Verification

You separated AI capability from exposure, human behavior, controls and impact — then built a response that still works when the fake is convincing.

✓ Recognize capability
✓ Assess exposure
✓ Choose stronger controls
✓ Verify independently
CHECK SOMETHING SUSPICIOUS →WHAT WOULD YOU DO IN A CYBER CRISIS? →