Mobile Edge AI Framework for Real-Time ESG Compliance Auditing through NLP and Blockchain Ledger Verification
Year: 2026
Type: article
Abstract: The Mobile Edge artificial intelligence (AI) framework is introduced as an environmental, social, and governance (ESG) verification system that integrates natural language processing (NLP) and permissioned blockchain to facilitate near-real-time ESG compliance auditing. A lightweight transformer model is employed to identify ESG-related claims and extract relevant indicators, numerical values, units, locations, and reporting times. Extracted claims are structured and recorded on the blockchain to ensure verifiable, tamper-resistant records. A discrepancy assessment mechanism then classifies each claim as supported, partially supported, contradictory, or unsupported. The prototype was evaluated using a synthetic Kaggle dataset containing 11,000 annual company records from 1,000 organizations across nine regions and multiple sectors. The ESG classification model achieved a true positive rate of approximately 93.2%, with F1-scores of 0.945, 0.915, and 0.895 for environmental, social, and governance claims, respectively. Claim verification rates ranged from 93.1% to 98.4%. Edge deployment reduced verification latency to 124 ms compared to 231 ms in the cloud environment, representing a 46.3% reduction. These findings indicate that the framework can enable proactive ESG auditing, enhance claim traceability, accelerate verification, and deliver explainable compliance-risk signals.
Institutions: Hasanuddin University, University of Baghdad
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $280
OpenAlex ID: W7223909672