VerityAI
Where AI Meets Accountability
Created on 18th October 2025
•
VerityAI
Where AI Meets Accountability
Description of your solution
VerityAI: Agentic News Verification Platform
Idea Overview
VerityAI is a news intelligence platform and browser extension powered by agentic AI, designed to continuously scan, detect, and verify information from diverse sources. It provides easy-to-understand, bias-aware, and factual updates, particularly during crises, empowering users to make informed decisions and counter misinformation in real time.
How It Works
- Continuous Multi-Source Scanning: Autonomous agents ingest breaking news and emergent claims from major media outlets, social feeds, and public datasets.
- Bias and Factuality Detection: Every source and article is assessed for bias (left, center, right) and factual reliability, visually displayed for user transparency.
- Automated Fact-Checking Pipeline: AI agents cross-reference claims with trusted databases, scientific literature, and verified government sources, flagging misinformation and highlighting evidence.
- Intuitive Browser Extension & Website: Users instantly see bias/factuality overlays on news stories and social media posts, with contextual summaries and audit trails.
Opportunities
- Reduces the spread and impact of misinformation during high-stakes event.
- Enhances public media literacy with interactive, transparent bias and factuality indicators.
- Personalizes crisis updates and trusted feeds for individuals, journalists, and decision-makers.
- Extends to other domains (health, finance, policy) due to scalable agentic AI architecture.
Differentiators
| Feature | VerityAI | Existing Platforms |
|---|---|---|
| Autonomous real-time agents | Yes | Mostly manual/static |
| Dynamic crisis prioritization | Yes | Rare |
| Visual bias/factuality breakdown | Advanced | Simple/limited |
| Claim-level verification | Agentic AI-driven | Often only at source |
| Transparent audit trails | Directly linked sources | Often opaque |
VerityAI stands out by proactively adapting to emergent events, providing real-time, claim-level verification, and delivering user-friendly context overlays in both web and social media environments.
Problem Solving
- Verifies claims and stories dynamically in fast-evolving crises, flagging both misinformation and uncertainty.
- Reduces blindspots by showing side-by-side bias splits and factuality ratings for every story, combating echo chambers.
- Gives users confidence in their news consumption with direct evidence and source transparency.
Unique Selling Proposition (USP)
VerityAI is the first autonomous, agentic AI platform to deliver real-time, bias-aware, fact-verified news overlays and context for every major event - empowering citizens, journalists, and leaders to respond decisively and accurately, especially in crises.
Technologies Used
- Agentic AI Frameworks: CrewAI, LangChain, Anthropic Claude agents.
- NLP & LLMs: GPT, BERT, Claude for claim extraction and semantic analysis.
- Bias/Factuality APIs: AllSides, MBFC, Ad Fontes Media, Google FactCheck API.
- Front-End & Extensions: React/Next.js for dashboard; Chrome/Firefox SDKs for overlays.
- Back-End: Python (FastAPI, Flask), streaming (Kafka), Cloud (AWS, GCP, Azure).
- Database: NoSQL for scalable storage, SQL for verdict tracking.
- Privacy: Local analysis for sensitive data, GDPR compliance.
This platform will empower users, journalists, and decision-makers to cut through information chaos, boosting both personal and public resilience in the age of frequent crises
Tracks Applied (1)
Misinformation: Create an Agentic AI system that continuously scans multiple sources of information, detects emerging misinformation, verifies facts, and provides easy-to-understand, contextual updates to the public during crises.
Cheer Project
Cheering for a project means supporting a project you like with as little as 0.0025 ETH. Right now, you can Cheer using ETH on Arbitrum, Optimism and Base.