VeriFact

Inspiration

With misinformation spreading faster than ever, we wanted to build something that helps people cut through the noise and actually trust what they read online. Every day, millions of users share unverified stories — sometimes without realizing it. VeriFact was born out of the idea that everyone deserves a quick, reliable way to verify information without needing to be a tech expert or journalist. VeriFact empowers users to make informed decisions about news consumption, develop critical thinking skills for media literacy, and save time by quickly identifying unreliable sources.

What it does

VeriFact is an AI-powered Chrome extension that instantly analyzes any news article for credibility, bias, and accuracy. With one click, it gives users a clear breakdown of how trustworthy a story is, detects political bias, separates facts from opinions, checks the author’s credibility, and flags outdated or misleading content.

• AI evaluates source reliability and content quality
• Identifies left/center/right political leanings
• Distinguishes factual claims from subjective
• Detects emotional manipulation and clickbait tactics
• Rates news outlets from A+ to Unknown
• Verifies journalist credentials and track record
• Flags outdated stories presented as current news
• Cross-references with reliable news organizations
• Summarizes main claims in bullet points
• Auto-translates foreign language articles
• Suggests search terms for cross-referencing
• Identifies suspicious content patterns

How we built it

We built VeriFact using Chrome’s Extension API with an integrated AI analysis pipeline powered by the Gemini Nano model for on-device inference. The extension combines multiple APIs (Prompt, Summarizer, and Translator APIs) for summarization, translation, and context understanding to evaluate news credibility in real time. We designed a clean, tab-based interface that organizes analysis results into Overview, Content, and Source sections. Everything runs locally, so no user data ever leaves the device, ensuring security and transparency.

Challenges we ran into

One major challenge we had was balancing accuracy and performance. We need to ensure that the extension could analyze articles quickly while still producing meaningful results. We also had to work around Chrome’s new privacy restrictions and optimize our use of the Gemini Nano model for efficient on-device processing.

Accomplishments that we're proud of

We’re proud that VeriFact runs entirely offline, providing instant, private, and reliable fact-checking right in the browser. We managed to combine multiple AI tools into one seamless experience, from sentiment analysis and fact-checking to translation and bias detection, all in a user-friendly interface.

What we learned

We learned how to use Gemini Prompt, Summarizer, and Translator APIs to build the extension. We learnt to build a clean and easy to understand layout to display the results and findings . We also discovered how on-device AI can dramatically improve both speed and privacy, setting a strong foundation for responsible AI development.

What's next for VeriFact

Next, we plan to expand VeriFact with community-driven fact-checking, letting users flag suspicious content and contribute to improving accuracy. We also want to build a mobile version and integrate real-time deepfake detection for images and videos. Ultimately, our goal is to make VeriFact a trusted companion across all digital platforms to help everyone think critically and navigate information with confidence.

Built With

• gemini-nano
• prompt-api
• summarizer-api
• translator-api