TruthTrace

TruthTrace – On-Chain Algorithm Dispute & Explanation Engine

🧠 Overview

TruthTrace is a full-stack, blockchain-powered platform that ensures transparency, accountability, and explainability in AI-driven decision-making systems. In an age where AI impacts critical aspects of human life—from finance and healthcare to criminal justice and hiring—TruthTrace addresses the fundamental challenge of trust in automated decisions.

🎯 Problem Statement

AI systems today operate as “black boxes.” Users often have no insight into how or why an AI model made a specific decision—whether it's approving a loan, denying insurance, or flagging content. This lack of visibility not only erodes user trust but also poses significant ethical, legal, and societal risks.

Moreover, even when users are affected by these decisions, there’s usually no standard mechanism to challenge or appeal them. This creates a critical gap in fairness, transparency, and user empowerment.

💡 Solution

TruthTrace bridges this gap by providing an end-to-end framework for logging, disputing, and explaining AI decisions through a combination of blockchain technology, community governance, and large language models (LLMs). It transforms opaque algorithms into auditable, explainable systems.

🔍 Key Features

🔐 Core Functionalities:

• AI Decision Logging: Every AI output is cryptographically hashed and recorded on-chain, ensuring immutability and tamper-proof traceability.

• Dispute Management System: A community-driven challenge mechanism allows users to dispute AI decisions. Validators vote to uphold or overturn outcomes.

• LLM-Powered Explanations: Each AI decision can be auto-explained in plain language using integrated large language models (e.g., GPT).

• Role-Based Access Control: Secure authentication with differentiated roles for users, admins, and validators.

• Blockchain Anchoring: Decisions, disputes, and explanations are permanently logged using smart contracts on the blockchain.

🌐 Frontend Experience:

• 3D Interactive Dashboard: An immersive UI built with CSS 3D transforms, animations, and Framer Motion to deliver real-time engagement.

• User-Centric Pages:

  • Dashboard: Personalized view of AI interactions

  • My Decisions: Track past AI decisions affecting the user

  • Dispute Center: File challenges and monitor resolution

  • Audit Explorer: Visualize blockchain-stored decision logs

  • Community Leaderboard: Validator activity and rewards

🏗️ Architecture

• Smart Contracts: Written in Solidity to manage logging, disputes, voting, and role control.

• Backend: Node.js & Express.js for API handling, MongoDB for off-chain storage.

• Frontend: React 18 with Material UI, Web3.js integration, and animated interactions.

• Blockchain Layer: Ethereum-compatible network for storing verifiable records.

• LLM Integration: Backend-connected APIs to fetch explanations for decision metadata.

⚙️ Technologies Used

• Blockchain: Ethereum, Solidity, Web3.js

• Backend: Node.js, Express.js, MongoDB, JWT

• Frontend: React, React Router, Material UI, Framer Motion

• AI: OpenAI GPT API for decision explanations

• Deployment: Smart contract deployment via scripts, local development via npm run dev

🚀 How It Works

1. Submit: An AI system makes a decision. The app submits this decision to the backend and blockchain.

2. Log: The decision metadata and hash are stored immutably via smart contracts.

3. Explain: The system fetches a human-readable explanation using LLMs.

4. Dispute: If a user challenges the decision, a dispute is logged and broadcast to validators.

5. Vote: Validators analyze the dispute and vote. The outcome is finalized and stored on-chain.

6. Audit: Every action is traceable via the Audit Explorer and open API logs.

✅ Use Cases

• Finance: Explain and dispute AI-based loan or credit decisions

• Healthcare: Audit diagnostic AI outcomes and get second opinions

• Recruitment: Validate fairness in resume shortlisting algorithms

• Legal Tech: Document AI-influenced parole or sentencing recommendations

• Content Moderation: Dispute wrongful bans or takedowns on social platforms

📈 Impact

• Promotes AI ethics, fairness, and trust

• Empowers end users to challenge potentially harmful AI outcomes

• Reduces the liability for companies using AI by offering explainability

• Provides a universal framework to bring transparency across all AI-driven systems