Can technology such as AI and data analytics strengthen legislative oversight of public expenditure? Discuss in the context of the CAG–PAC relationship.

The CAG audits government accounts. The Public Accounts Committee (PAC), India's oldest parliamentary committee (set up in 1921), uses those audits to hold the executive to account [3]. Some PAC reports have now been pending for 15–20 years. In November 2026, Speaker Om Birla called for AI and data analytics to be used in this work [1]. Technology can strengthen this chain a great deal, but only if procedures are also reformed.

CAG audit → Report laid in House → PAC selects paras → Ministry ATNs → PAC report
        ↑ AI: risk-flagging        ↑ AI: prioritisation   ↑ AI: tracking/alerts

Fig: Where technology fits in the CAG–PAC chain

How technology can strengthen oversight

• Choosing what to examine: Anomaly detection across large volumes of spending data can flag high-risk spending. This helps the CAG target its audits and the PAC choose which audit paragraphs to hear. It also helps answer how PACs should treat CAG reports [1].
• One shared record: In 2013 the Finance Ministry counted 815 pending Action Taken Notes (ATNs) while the CAG counted 722 [4]. A digital tracker shared by the ministry, the CAG and the PAC would end such gaps.
• Tracking replies: Ministries must submit ATNs within three months of a report being laid, and the CAG checks them [5]. Dashboards with automatic alerts can show which ministries are late.
• Keeping inquiries on track: Lok Sabha members of the PAC are elected every year [3]. Searchable digital case files help new members take over long-running inquiries.

Limits of technology

• Ministries that do not reply: The backlog lies mainly with ministries that send ATNs late. The four-month deadline set in the PAC's 105th Report (1995-96) is still missed [4]. Software can show a delay but cannot force a reply.
• Advisory, after-the-event role: PAC findings are recommendations only and come after the money is spent [2]. Analytics cannot make them binding.
• A few departments drive most of the backlog: Railways, Defence and the tax boards held a large share of pending paragraphs [4]. Clearing them needs political pressure, not just algorithms.
• Capacity gaps: Analytics need clean, linked accounts data and trained staff in legislature secretariats.

Technology can multiply the impact of the CAG–PAC partnership but cannot replace it. AI-enabled tracking should be combined with ministry-wise public reports of overdue ATNs, enforced deadlines and floor debate on PAC reports. Together these would make the PAC an effective "watchdog of the watchdog" and strengthen the executive's accountability to the legislature under Article 75(3).