Jesia Unveils Real-Time Verification Technology for AI Decision-Making

"Building a Data Sovereignty Trust Infrastructure"

Jesia Platform announced on August 11 that it has unveiled a circuit design technology based on 'Zero-Knowledge Proof (ZKP)' that enables real-time verification of the entire decision-making process made by AI.

This technology was recently registered as a draft in Ethereum EIP-7911. According to the company, this marks the first time a Korean company has submitted AI decision verification technology for discussion as an international blockchain standard.

Zero-Knowledge Proof is a cryptographic technique that proves a certain claim is true without revealing the underlying data or internal logic that supports the claim. For example, it can prove only the 'fact' that AI reached a specific conclusion, without exposing the original data or detailed calculation process used in that decision. This allows for the protection of personal information and industrial secrets while still ensuring the reliability of the decision.

Jesia's ZKP circuit consists of seven stages. Data collected from IoT-based systems in industrial settings is encrypted through a ZKP generator, and an AI agent makes decisions according to policy conditions. Subsequently, a large language model (LLM) interpreter and an exception condition analyzer refine the context of the decision and generate a 'pattern verification tag.' This tag is automatically compared in real time with industry-specific standards, policy conditions, and ethical principles to prove the legitimacy of the decision.

The ZKPs and decision tags generated throughout the entire process are stored on an Ethereum L1 rollup-based blockchain, making the decision history tamper-proof. This enables the implementation of a networked trust structure where multiple AI agents can mutually verify each other's decisions.

The ZKP pattern verifier developed independently by Jesia automatically analyzes the flow of AI decisions according to industry and policy standards, and verifies their validity in real time without violating privacy. The company explained that this allows both enterprises and governments to secure the reliability and utility value of their data simultaneously.

Jesia plans to expand this technology into an ▲AI decision audit system ▲automated verification environment among agents ▲industry-specific certification response modules, applying it to ESG reporting, GX (Green Transformation) compliance, industrial security, and international certification.

A Jesia official stated, "What matters most is proving the entire AI decision-making process in a trustworthy structure," adding, "This technology will serve as a foundation for Korea to become a leading nation in digital trust."

This content was produced with the assistance of AI translation services.