CYBERDIVA LABS
AI RISK LEARNING UNIT #001 · THE CALL

YOUR DAUGHTER CALLS.
SOMETHING IS WRONG.

A familiar voice. An urgent request. You have seconds to decide what you trust.

Sound on recommended.

01 · EXPERIENCE
SCENE 1 / 4

The story will continue automatically.

DECISION POINT

THE VOICE SOUNDS REAL.
THE EMERGENCY FEELS REAL.

You have seconds to decide.

What would you do?

WHAT HAPPENED?
02 · DECONSTRUCT

The voice was only one piece.

The attack became convincing because technology + context + pressure worked together.

🤖 AI IMITATIONA voice model can imitate vocal characteristics from available audio.
🌐 BELIEVABLE CONTEXTPublic posts can reveal relationships, names and what someone is doing.
⏱️ PRESSUREFear and urgency make it harder to stop and verify.
Key idea: AI did not create the whole attack. It made one part — the voice — more convincing.
03 · ASSESS

What actually breaks the attack?

Choose ONE answer.

04 · PROTECT

Build your defense

Now choose ALL the controls that still work even when the fake is convincing. Tap again to deselect.

05 · BUILD YOUR PLAN

Create your family verification rule

If a family member suddenly calls asking for money, what will your family do?

1
PICK A PRIVATE PHRASE

Think of it offline — don't type it here.

✓ Keep it private and not guessable from social media.
2
WHEN SHOULD YOU USE IT?

Choose the strongest rule.

3
WHAT COMES NEXT?

A correct family phrase is helpful. What should you still do?

06 · PRACTICE

Try the call again

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

This time, build the safest response one step at a time.

STEP 1 · INTERRUPT THE PRESSURE
STEP 2 · CHANGE THE CHANNEL
STEP 3 · ADD YOUR FAMILY CONTROL
SKILL UNLOCKED

VERIFY THE PERSON.
NOT THE VOICE.

A convincing voice is not proof of identity. Your strongest move is to leave the suspicious channel and verify independently.

EXPERIENCEFeel how a believable AI-enabled scenario creates pressure.
DECONSTRUCTSeparate voice imitation, public exposure and human pressure.
ASSESSDecide what you can — and cannot — trust.
PROTECTChoose controls that don't depend on the suspicious caller.
BUILDCreate a family verification rule.
PRACTICEPause → leave the channel → verify independently.
YOUR RULE

PAUSE → LEAVE THE CHANNEL → VERIFY INDEPENDENTLY

CHECK SOMETHING SUSPICIOUS → EXPLORE ANOTHER SCENARIO →