WonderAI - AI Misinformation Detector

Verify Before You Trust With Wonder AI

Comment

Inspiration

In this era, information can spread across the globe instantly. However, the problem is that many people accept this information at face value and believe it immediately, even when it is misinformation. This often happens because people might be reluctant to fact-check or are overwhelmed by too many conflicting perspectives. Therefore, this project was created as a tool for those who feel doubtful about the information they receive, or for those who are simply curious and want to learn more about other instances of misinformation. I aim to build a "digital shield" that empowers users to think critically with our motto: "Verify Before You Trust."

What it does

WonderAI is an intelligent platform designed to combat hoaxes through two primary pillars:

  • Real-time AI Analysis: Users can paste article text or URL links. Our AI engine, WonderAI, analyzes the input to detect logical fallacies, sarcasm, and factual inconsistencies.
  • Global Trending Monitor: A dashboard that displays live, verified fact-checks from global publishers via the integration of the Google Fact Check Tools API.
  • Transparency First: The application provides a Confidence Score for every analysis and allows users to automatically download a PDF of AI ethics and guidelines to understand how the machine makes its decisions.

How I built it

Iutilized a high-performance technology stack to ensure a premium user experience:

  • Frontend: Built using React (Vite) and Tailwind CSS v4 to create a modern "Glassmorphism" interface. I implemented Framer Motion for complex layout animations, including a custom splash screen and smooth shared-element transitions.
  • Backend: A Node.js and Express server handles the API logic and data communication.
  • AI Engine: Powered by Google Gemini 2.5 Flash. I implemented advanced prompt engineering so the AI can automatically detect the input language and provide responses in specific categories: 'Likely Misinformation', 'Questionable', 'Likely Factual', or 'Satire'.
  • Data Integration: Integrated the Google Fact Check Tools API to provide external, human-verified data alongside AI reasoning.

Challenges i ran into

  • API Rate Limiting: AI models have specific usage quotas. I built robust auto-retry logic (up to 3 attempts) and a user-friendly error handling system to manage "429 Too Many Requests" errors gracefully.
  • The "Flicker" Effect: Creating a seamless transition from the splash screen logo to the navbar logo was a technical challenge. I solved this using Framer Motion's layoutId to animate the logo consistently across different components.
  • Source Credibility: Analyzing live URLs without scraping every site required us to train the AI to evaluate domain reputations and URL patterns when live content was inaccessible.

Accomplishments that i'm proud of

  • WonderAI Assistant: Successfully creating an AI persona that remains neutral, objective, and supports multiple languages.
  • Smooth UX: Achieving an interface performance that feels like a native app, especially with the Bounce Animation on the logo and the Moving Gradient background.
  • Transparency Mechanism: The ability for users to generate a professional PDF report of our ethics guidelines directly from the browser using the jsPDF library.

What i learned

  • Prompt Engineering: I learned that the quality of AI analysis depends heavily on the precision of the system instructions provided.
  • Ethical AI: Building this tool taught us the importance of transparency. I learned that AI should be positioned as a "second opinion," not an absolute truth that cannot be contested.
  • Asynchronous Flow: Managing parallel API calls between the AI and the Fact Check API while keeping the interface response time fast.

What's next for WonderAI - AI Misinformation Detector

  • Browser Extension: Creating a Chrome extension that automatically highlights potential misinformation while users browse social media.
  • Community Verification: Adding a community-based verification layer where users can contribute to validating information.
  • Image/Video Deepfake Detection: Expanding the engine's capabilities to analyze images and videos for AI manipulations.
  • Mobile App: Bringing WonderAI to iOS and Android platforms for information verification anywhere and anytime.
// Our core philosophy in code:
const wonderAI = {
  mission: "Restore trust in information",
  engine: "Server (Gemini 2.5 Flash)",
  motto: "Verify Before You Trust"
};

Built With

Share this project:

Updates