AI saved my parents from online scams

My aging family kept running into scams.

My father clicked on a fake insurance email and got his computer hacked. My mother-in-law nearly wired $5,000 to someone pretending to be her grandson after receiving a call claiming he had been arrested in Mexico.

She was literally at the bank filling out the wire transfer paperwork when a teller gently asked, “Is everything okay?”

When she explained what was happening, the teller immediately recognized it as a scam and had her call her grandson directly. He was right down the road, completely fine.

That moment stuck with me.

I work in banking and have spent years around fraud prevention and cybersecurity training, so I could usually spot scams quickly. But I realized most people do not have that background.

And honestly, almost everyone has a story like this now.

It is not just seniors. One coworker told me about a younger relative in her twenties who got a call from someone pretending to be her bank and started giving away sensitive information over the phone.

The problem was not intelligence. It was how easily fear, urgency, and trust can override caution in the moment.

I decided to build a simple mobile app to help people recognize scams and slow down before reacting.

I had never really built a mobile app before, so I started using AI tools like ChatGPT, Perplexity, and Cursor to help me create it.

At first, I explored ideas like intercepting scam calls directly, but the technical complexity became overwhelming. AI helped me rethink the problem and simplify the approach.

Instead of trying to stop scams technologically, I focused on awareness and behavior.

I used Reddit discussions and real scam examples as research material, then asked AI to identify common manipulation patterns, warning signs, and situations people should be skeptical of.

I also used AI to help make the app accessible and non-threatening, especially for older users. Large fonts, simple navigation, and clear language mattered more than advanced features.

One thing that was important to me was privacy. I intentionally designed the app not to collect user data.

It would be pretty ironic if the app teaching people to protect their information was also collecting it.

My parents became noticeably more aware of scams.

Now when we talk each week, they tell me about suspicious emails and scam attempts they recognized instead of reacting to them automatically.

One recent example was a fake message claiming there was a Driver’s License issue involving the state police. Instead of panicking, they were skeptical immediately.

The biggest shift was not that the app could stop every scam. It helped create a pause between fear and reaction.

Instead of responding emotionally or urgently, my family became more thoughtful about what deserved trust.

What started as a side project turned into a new form of caregiving.

Step 1: Start with a real-world problem you deeply understand The project came from seeing family members repeatedly targeted by scams and realizing most people are not trained to recognize manipulation tactics

Step 2: Use AI tools to reduce the technical barrier to building AI tools like ChatGPT, Cursor, and Perplexity helped turn a non-mobile developer into someone capable of building and iterating on an app idea

Step 3: Use real scam stories as source material Reddit discussions, news stories, and personal experiences helped identify common patterns like fake bank fraud calls, Medicare scams, and urgent family emergencies

Step 4: Let AI help simplify the solution The original idea involved intercepting scam calls directly, but AI helped rethink the problem and focus instead on awareness and behavior change

Step 5: Design around trust and accessibility The app prioritized large fonts, simple navigation, and minimal friction so it would feel approachable for older users

Step 6: Align the product with the mission The app intentionally collects no personal user data because the goal was to teach people how to protect their information, not harvest it