Integrating Blockchain and Ontologies in Artificial Intelligence Regulation

Artificial intelligence has evolved to possess advanced cognitive abilities, leading to a heightened awareness of ethical and legal dilemmas and potential threats. The training of AI models involves processing data that may violate established legal and moral norms, prompting widespread concern. It is essential to closely monitor and document the deployment of AI models to prevent illicit activities. Despite the establishment of various standardized frameworks for AI governance, achieving transparency and uniformity remains a persistent challenge. This study proposes a novel framework that combines blockchain technology with ontologies to provide a regulation-approval record showing that AI models have passed compliance auditing. The decentralized nature of blockchain ensures no authority over the trained models. Moreover, developing a concise ontology to save AI models in compliance will reduce the cost of keeping the data on the blockchain. Adherence to this framework provides a roadmap for ensuring transparency, semantic traceability, and unified standards of compliance with international ethical and legal standards.