1. The Marketing Campaign That Knew Too Much — and Worked

I recently spoke with someone at a financial services firm that's deeply integrating AI with their marketing stack. Their goal? Create hyper-targeted customer outreach at scale.

They had transaction-level visibility into customers’ spending patterns, knew where they lived, what they bought, where they traveled. Then, they layered in public and third-party data — things like neighborhood demographics and product trends.

Using AI, they built a multi-month personalized campaign across 30,000+ customers. Every email was hyper-personalized — specific to that person’s habits, lifestyle, and even the local buying trends in their neighborhood.

One example: if your neighbors recently bought a specific truck model and your online behavior suggested automotive interest, you received an email that nudged you — subtly but deliberately — toward that very type of vehicle.

It was brilliant. And also... a bit unsettling.

2. When Familiar Faces Aren’t Really Familiar

About two weeks ago, my wife’s WhatsApp account was hacked. It started with a seemingly innocent code request from a known group contact — which she responded to, unknowingly transferring her account to the hacker’s device.

The real story started after.

The hacker didn’t just spam her contact list — they took their time. They monitored ongoing chats. They interacted like her — liking comments in family chats, holding brief conversations, and of course, asking for emergency money from a friend. (And someone did transfer money, and recovered it - but that's story for another day!)

One friend, suspicious, asked why she was online so late. The reply? “Up because of this issue, but can’t talk right now. Please send the money.” When another person tried to warn the group that her account was compromised, the hacker deleted the message.

It was disturbing — not just because of what happened, but because of what could happen as AI advances.

Imagine this same scam with AI usage:

• An AI trained on her chat history starts responding in her tone and style.

• Voice synthesis technology creates perfectly mimicked phone conversations.

• The scammer no longer needs to be clever — just connected to an AI model.

Case 1: When AI Fools Diplomats

This isn’t hypothetical: remember how an unknown individual impersonated Secretary of State Marco Rubio using AI-generated voice and text messages.

The impostor contacted at least five high-level officials — including three foreign ministers, a U.S. governor, and a member of Congress — using AI to replicate Rubio’s voice and communication style. This wasn't a phishing email — it was AI-driven diplomatic deception.

Source: The Washington Post, July 8, 2025.

Case 2: Deepfake Scams Are Going Global

In India, these scams have become dangerously common. A retired government employee in Kerala was duped out of ₹40,000 via an AI-generated deepfake video that mimicked a friend requesting urgent help.

Worse still, in Delhi, a senior citizen was targeted with a voice-cloned message — allegedly from a kidnapped child relative — begging for help. In panic, he transferred ₹50,000 before realizing the child was safe.

(Source: NDTV.com, AI Scams Surge, October 2024)

The Data Doesn’t Lie: Scams Are Growing — Fast

According to Revolut’s Consumer Security and Financial Crime Report (H2’24):

• Meta platforms (Facebook, WhatsApp, Instagram) accounted for 54% of all authorized fraud cases.

• WhatsApp fraud rose by 67% between H1 and H2 2024.

• Telegram fraud grew by a staggering 121% in the same period.

This isn’t just a tech issue — it’s a collective failure in how we manage digital identity, AI misuse, and trust online.

Walking the Line: A Call for Thoughtful Innovation

The same data and AI that help us serve customers better can also be weaponized. For those of us in financial services, we’re often both the enablers and defenders of this line.

This isn’t just about cybersecurity. It’s about trust. And trust, once lost, doesn’t come back easily.

AI may personalize our outreach, but our responsibility to protect people must remain universal. Let’s build systems that are not only smart — but safe.

These two events — one from a boardroom, one from my home — brought the same message into sharp focus: the line between helpful and harmful is getting thinner. As leaders in our respective areas, we need to sharpen our understanding and strengthen our guardrails.