Building a scam awareness app: what goes into one (from the team behind SCAMalicious)
Scammers are annoyingly good at their jobs. They're patient, they test what works, and they update their scripts faster than most of us update our phones.
So when a client asked us to build SCAMalicious, an app to help people spot and report scams, the interesting question wasn't "can we build a mobile app?" It was "how do you build something that's actually useful in the thirty seconds between a suspicious text arriving and someone clicking the link?"
The app is now live on Google Play with more than 5,000 downloads, and the App Store version is in review. Here's what goes into scam detection app development, and the decisions that mattered. (We built the mobile app. The client's website and brand are theirs.)
First, be honest about what "detection" means
Let's get the uncomfortable bit out of the way. A consumer app can't see inside every scammer's head, and any app that promises "100% scam detection" is, ironically, making a slightly scammy claim.
What an app can do is help a person check a situation against the warning signs. That reframing shaped everything else in SCAMalicious. The core feature isn't a magic scanner. It's a guided check that gets the user to slow down and look at what's in front of them.
The risk check: a few questions, an instant answer
The heart of the app is the scam-risk check. The user picks the kind of scam they're worried about (a dodgy online shop, a prize that needs a "small fee", a call from "the bank"), answers a few plain questions about what happened, and gets an instant result on how likely it is to be a scam.
Three design choices made this work:
- Few questions, plain words. Someone panicking about a message won't fill in a long form. Short, clear, done.
- Instant result. No waiting, no "we'll email you". The moment they need an answer is right now.
- Saved history. Every check goes into the user's history, so they can look back, or show a worried relative what the app said.
Teaching the warning signs
A result is useful once. Knowing the signs is useful forever. So alongside the check, the app has a learning centre and warning-signs guides for common scam types.
The aim is that after a few weeks the user doesn't need the app for the obvious ones. That might sound like a strange goal for an app. It's the right one for a safety app, and users can tell the difference between an app that wants to help them and one that wants their attention.
Reporting, country by country
Here's a problem most people don't think about until it happens: you've been scammed, and you have no idea who to tell. Police? Your bank? Some government fraud office with a website from 2009?
SCAMalicious handles this with country-based reporting. Pick your country and the app shows the right fraud authorities to contact. It sounds simple. The work is in the data: keeping a list of organisations per country accurate and useful is ongoing content work, not a one-time feature. We'd tell anyone building a similar app to budget for that from the start.
Community stories (and moderating them)
Scammers thrive on embarrassment. People who've been caught often don't tell anyone, which is exactly what lets the same trick work on the next person.
The app's anti-scam community gives people a place to share what happened to them and what they learned. That's powerful, and it comes with responsibilities: user content needs guidelines and moderation, so the community stays helpful and doesn't become a new place for scammers to post. If you're building something similar, design the moderation flow before launch day, not after the first bad post.
The AI assistant
Some questions don't fit a questionnaire. "My landlord wants the deposit in gift cards, is that normal?" (It is not.) For those, SCAMalicious has an AI assistant powered by OpenAI and Grok models, served through custom APIs.
The important part isn't the model, it's the rules around it. A safety assistant has to stay on topic, help the user think the situation through, and point to proper reporting rather than play detective. We wrote more about this in adding an AI chat assistant to a mobile app.
Paying for it: RevenueCat
AI and good content cost money to run, and keeping the core safety tools free matters for an app like this. Subscriptions and in-app purchases cover the premium side, handled with RevenueCat so one setup works across Google Play and the App Store. Building two billing systems for two stores is a special kind of pain we'd recommend avoiding.
Why Flutter and FlutterFlow fit this app
A scam awareness app is mostly accounts, forms, content and lists, with a few heavier pieces (AI, payments, server logic) on top. That's an ideal shape for FlutterFlow: the standard screens come together quickly, and the specialised parts go into custom Flutter code and the backend, which here is Firebase with Cloud Functions plus custom APIs.
One codebase also means iOS and Android get the same app. The Android version went live first and the iOS version is following through Apple's review, built from the same code. If you're curious how that tradeoff plays out more generally, we compared FlutterFlow vs native development here.
If you're building a safety app
A few things we'd tell anyone starting one:
- Don't promise certainty. Promise help.
- Make the first action fast. Seconds count when someone is about to pay.
- Budget for content, especially reporting data per country. It's never "done".
- Moderate community content from day one.
- Keep AI behind your own backend, with clear rules on what it will and won't say.
You can read the full SCAMalicious case study to see how the pieces fit together.
Building an app that protects people, whether that's from scams, fraud or something else? Tell us what it needs to do. We're a certified FlutterFlow team and we'll tell you straight which parts are quick and which ones deserve the extra care.
A few things people ask us
How does a scam detection app work?
Most consumer scam apps don't magically know whether a message is a scam. They help the user check: pick the type of situation, answer a few questions about what happened, and get a risk result based on known warning signs. Good ones also teach the signs, point to the right place to report, and let people learn from others' experiences.
Can an app tell me for certain that something is a scam?
No, and an honest one won't claim to. It can tell you how many warning signs a situation has and how likely a scam is. The final call is yours, which is why clear guidance and reporting options matter as much as the result.
What features should an anti-scam app have?
A quick risk check, guides to common scam types and their warning signs, a way to report scams to the right authority in your country, a history of your past checks, and some form of community or stories. An AI assistant for follow-up questions is a strong addition if it's kept on topic.
Is FlutterFlow good for building an app like this?
Yes. A scam awareness app is mostly forms, content, accounts and lists, which FlutterFlow builds quickly, and the parts that need more control (AI, payments, server logic) can be handled in custom Flutter code and a backend. That's how we built SCAMalicious.
How do anti-scam apps make money?
Commonly through subscriptions or in-app purchases for premium features, while keeping core safety tools free. RevenueCat is a popular way to run subscriptions across both app stores from one setup.