खंड 3 No. issue2,April2026 ,pg2208-2223 (2026): Explainable AI Assisted Smart Contract Auditing for Administrative Workflows

Smart contracts can encode administrative approvals, payments and record transitions, yet a valid transaction may still breach an institution’s intended procedure. This exploratory paper proposes an explainable AI-assisted auditing framework that separates vulnerability detection, explanation, repair generation and independent repair verification. A focused review of primary research and official documentation informs a workflow specification, threat model and reproducible evaluation protocol. Static analysis and transaction-oriented testing supply program evidence, while AI helps relate code to approved administrative requirements. Explanations identify the affected rule, source statements, state conditions and supporting evidence. Candidate repairs are assessed against frozen specifications, protected regression tests and applicable formal properties. An illustrative payment case shows why preventing a prohibited action is insufficient when the same change blocks legitimate activity. The proposed evaluation compares conventional auditing, AI assistance, evidence-linked explanations and independent verification under matched budgets. Detection accuracy, explanation fidelity, repair acceptance, workflow preservation and cost are measured separately. The contribution is a requirements-centred design and evaluation protocol; no prototype results or institutional field observations are claimed.

Keywords:

Explainable artificial intelligence; smart-contract auditing; administrative workflows; vulnerability detection; automated repair; independent verification.