Integrating Large Language Models and AI Into Blockchain: A Framework for Intelligent Smart Contracts and Fraud Detection

Abstract

The convergence of Artificial Intelligence (AI) and Large Language Models (LLMs) with blockchain technology is transforming information systems by enhancing their efficiency, security, and decision-making capabilities. This research explores the integration of AI and LLMs, such as GPT and BERT, into blockchain-based information systems to address challenges related to data integrity, transaction processing, and smart contract automation. A layered architecture is proposed, comprising an AI-powered query engine, an LLM-enhanced decision-making layer, and a data synchronisation module bridging on-chain and off-chain environments. The system evaluation highlights significant improvements in transaction efficiency, query accuracy, energy savings, and security threat detection rates. Experimental results demonstrate a 32% reduction in transaction latency, a 20.5% increase in fraud detection accuracy, and a 23% reduction in energy consumption. These findings underscore the viability of integrating AI and LLMs with blockchain technology for developing intelligent, secure, and scalable information systems.