The central government has initiated a plan to deploy artificial intelligence to scrutinise nearly 2.9 lakh pending claims for unclaimed shares, unpaid dividends, and matured deposits. The technological intervention is designed to identify forged documentation, impersonation, and emerging fraudulent patterns in refund submissions through 12 dedicated AI-based services. Operating under the Ministry of Corporate Affairs, the Investor Education and Protection Fund Authority currently relies on largely manual evaluations to process applications. The technical architecture for the overhaul is being established by the National e-Governance Division under the Ministry of Electronics and Information Technology to verify claimant credentials and flag suspicious applications before funds are released.
Targeting Document Manipulation in Deceased Investor Claims
The official request-for-quotation framework lists approximately 28 identified and projected fraud typologies encountered in investor claims. Key concerns focus on altered entitlement certificates, manipulated financial figures, and duplicate claims filed against identical shareholdings. Significant vulnerabilities persist in claims involving deceased investors, where bad actors frequently submit forged death certificates and falsified signatures on no-objection documents. The automated services will actively screen for deliberate efforts to conceal rightful legal heirs entitled to shared proceeds, ensuring that legitimate beneficiaries are not unlawfully deprived of inherited holdings.
Detection of Shared Data Identifiers Across Multiple Filings
The analytical framework will also track shared contact points and financial pathways that suggest coordinated irregularities. Algorithms are set to identify instances where single mobile numbers link to multiple Permanent Account Numbers, or where matching email addresses and identical bank identification codes are tied to unconnected applicants. Although shared identifiers will not serve as definitive proof of wrongdoing, they will provide critical warnings for manual follow-up. The system will similarly flag bulk volumes of templated application justifications that display structural traits typical of large language models, prompting officials to evaluate the underlying paperwork more rigorously.
Human Oversight Retained with Four-Month Rollout Goal
Despite the introduction of cross-database verification and automated pattern matching, final administrative authority will remain entirely with government personnel. The platform is engineered to generate claim summaries and propose internal processing codes, but it will lack the statutory power to approve, reject, or formally declare an application fraudulent. Departmental officers must independently weigh risk scores against primary evidence to prevent genuine applicants from facing adverse decisions caused by false alerts. The project documentation stipulates a four-month deployment schedule, supported by a baseline 12-month maintenance phase that can be extended by two additional one-year blocks. Following the tender issuance on October 13, 2026, bidding is scheduled to close on November 5, 2026.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
