AI Audit Agents are autonomous software systems that execute substantive audit and assurance tasks by applying large language models and reasoning algorithms to planning, evidence gathering, control testing, and workpaper documentation. These agents serve external audit firms conducting financial statement audits as well as internal audit and SOX compliance teams. Common capabilities span risk assessment automation, transaction sampling and testing, journal entry analysis, contract review, and narrative workpaper generation. The category emerged from traditional audit analytics and robotic process automation but has rapidly evolved since 2023 with the introduction of LLM-powered agents capable of interpreting unstructured evidence, drafting audit procedures, and performing multi-step reasoning across financial data sources. Deployment models range from task-specific copilots to fully autonomous agents that complete entire audit procedures with human review.