Fraudsters begin by collecting personal information from:

• Social media profiles

• Public videos or voice clips

• Data leaks and breaches

• Online purchases, comments, and posts

AI tools are then used to analyse this data and build a detailed victim profile, including relationships, language preference, profession, and emotional triggers.

Next, AI is used to create credible impersonations, such as:

• Cloned voices of family members or officials

• Deepfake videos of police, executives, or public figures

• Fake documents (FIRs, notices, bank letters)

• Official-looking websites or backgrounds

This step establishes trust and legitimacy.

Fraudsters deliberately trigger emotions such as:

• Fear (arrest, legal trouble, account suspension) • Panic (kidnapping, medical emergency)

• Greed (guaranteed returns, exclusive offers)

• Authority pressure (“official instruction”, “confidential matter”)

AI enables these messages to be precise, realistic, and personalised, reducing rational thinking.

Victims are told:

• “Act immediately”

• “Do not disconnect”

• “Do not inform anyone”

• “This is confidential”

This isolates the victim from verification and support, a critical stage in most AI-enabled frauds.

Once trust and panic are established, fraudsters push for:

• Immediate fund transfer (UPI, bank, crypto)

• Sharing of OTPs, passwords, or KYC details

• Clicking malicious links

• Installing remote access or fake apps

AI bots or scripts may guide victims step-by-step during this phase.

After execution:

• Communication is cut off

• Numbers/accounts are abandoned

• The same AI voice, video, or script is reused to target new victims

This allows fraudsters to scale operations rapidly with minimal effort.

AI frauds succeed by combining technology with psychology. Pausing, verifying through trusted channels, and questioning urgency are critical to breaking this fraud cycle.