CAG to launch IIT Madras-built AI assistant for auditors

The Comptroller and Auditor General of India (CAG) plans to introduce an in-house language model on Monday to support auditors with queries, calculations and preliminary analysis. Developed exclusively for CAG by IIT Madras, LLM 2.0 has been trained on around 65,000 internal inspection reports, alongside auditing standards and guidelines. The platform will complement CAG’s conventional audit management system.

Deputy comptroller and auditor general Krishnan Sangaran Subramanian said the system would handle routine numerical work and highlight patterns in financial information. It may include separate streams for compliance, financial and works audits. Auditors will only be able to ask questions based on the underlying reports used to train the model and will not be allowed to upload material to the platform or the cloud. The tool will serve as an assistant rather than conduct audits independently.

CAG is also shifting its audit planning from individual units toward patterns identified across departmental data for 2026-27. Teams are examining digital benefit transfers, public works and salary payments for red flags. Integrated audits covering social schemes, ease of doing business, ease of living, multimodal transport and foundational literacy are underway, with initial reports nearing completion and expected to be presented soon.