Factify AI

“Detect truth. Defeat fake.”

Comment

Inspiration :The idea for Factify AI came from noticing how quickly fake news spreads online, especially on social media. People often share information without verifying it, and it becomes difficult to separate truth from misinformation. I wanted to build something simple and useful that could help anyone instantly check the authenticity of news and understand the reasoning behind it, instead of blindly trusting or rejecting it.

What it does :

Factify AI is a web application that helps users quickly determine whether a piece of news is real or fake. Users can simply paste any news text or link, and the system analyzes it in real time to provide a clear verdict—FAKE, REAL, or UNCERTAIN—along with a confidence score and explanation.

Powered by the Gemini API by Google, the platform goes beyond basic detection by understanding context, tone, and credibility. It also highlights suspicious sentences, evaluates emotional bias, and presents results through an interactive dashboard, making it easy for users to understand why a piece of information may be misleading.

How we built it :

I built Factify AI as a full-stack web application. The frontend was developed using HTML, CSS, and JavaScript to create a clean and interactive dashboard. For the backend, I used Node.js and Express to handle requests and securely connect to the AI service. The core intelligence comes from the Gemini API by Google, which analyzes the news content and returns structured insights like verdict, confidence, and explanation.

The workflow can be represented as:

User Input→Backend Server→Gemini API→AI Analysis→Frontend Dashboard

Challenges we ran into :

One of the biggest challenges was moving from a simple frontend-based simulation to a real AI-powered system. Setting up the backend and integrating the Gemini API by Google required handling asynchronous API calls, managing errors, and ensuring smooth communication between frontend and backend. We also had to make sure the API key stayed secure using environment variables. Another challenge was converting the raw AI response into structured data and mapping it correctly to different UI components like the dashboard, charts, and highlights.

Accomplishments that we're proud of

We’re proud of building a fully functional AI-powered fake news detection system from scratch. Successfully integrating real-time analysis using Gemini API and presenting it through a clean, interactive dashboard was a major achievement. The ability to not just detect fake news but also explain the reasoning behind it makes the project more meaningful and user-friendly.

What we learned

Through this project, we learned how to build a complete full-stack application and integrate AI into a real-world use case. We improved our understanding of API handling, backend development, and secure key management. We also learned how important prompt design is for getting accurate AI responses, and how to present complex data in a simple and understandable way for users.

What's next for Factify AI

In the future, we plan to improve the accuracy by combining multiple data sources and adding real-time fact-checking from trusted platforms. We also want to introduce multilingual support, a browser extension for quick checks, and a chatbot feature for interactive queries. Our goal is to make Factify AI more powerful, accessible, and widely usable for everyday fact-checking.

Share this project:

Updates