Fake Check
AI-powered protection for your conversations.
Created on 13th April 2025
•
Fake Check
AI-powered protection for your conversations.
The problem Fake Check solves
Deepfakes aren’t just cool tech tricks anymore — they’ve become a serious problem.
From fake voice calls to edited images and AI-generated text, people are getting tricked, scammed, and misled without even realizing it. Someone could call you sounding like your friend, send you a fake image that looks real, or share content that feels human-written but was entirely AI-generated. Scary, right?
That’s where our project steps in.
We're building a system that can detect deepfakes across different formats — whether it's a phone call, a video, an image, or even a block of text. Our goal is to protect users from fraud, scams, and misinformation by flagging anything that seems artificially manipulated.
Challenges we ran into
One of the major roadblocks we faced was with the Android platform itself.
Android doesn’t provide a native Call Recording API anymore due to privacy and security restrictions. This made it extremely difficult to directly capture the voice audio during live phone calls — which is a critical part of detecting fake or spoofed voices.
What this meant for us:
We had to explore alternative methods like using accessibility services, VoIP call recording, and microphone-based recording — all of which came with their own limitations.
Some of these methods were device-specific or worked only on certain Android versions.
Technologies used