Veritas - Fake News Detector AI

Inspiration

It is difficult to identify fake news, especially with new technologies that can generate content (images, text, etc.). We aim to create a product that assists users in verifying information with just a mouse click.

What it does

The tool scans an article or a user’s text input, analyzes it, and returns a confidence score to help determine its credibility.

How we built it

Back-end:

• Used Python to train an AI model to analyze article content and identify patterns associated with misinformation.
• Integrated the model into a Flask-based API for communication with the Front-end.

Front-end:

• Created a Chrome Extension using Javascript, HTML, and CSS.
• Utilized the Chrome API to read and manipulate the current tab DOM.

Challenges we ran into

• Gathering sufficient data to train the model.
• Ensuring seamless communication between the backend and the Chrome extension.
• Fine-tuning the model for high accuracy.

Accomplishments that we’re proud of

• Building an AI model and a minimum viable product (MVP) in just 3 days.
• Effective collaboration and teamwork.

What we learned

• The importance of communication and teamwork.
• How to rapidly develop and deliver an MVP to validate ideas and functionalities early on.

What’s next for Veritas

• Collecting more data to refine our AI model.
• Enhancing the UI/UX for a more user-friendly experience.

Built With

• css
• flask
• html
• javascript
• python