Two vulnerabilities in the AI-driven debt collection platform CreditNirvana have exposed 6.6 million accounts and approximately 315,000 call recordings, according to research released on Wednesday. The platform, owned by Indian fintech company Perfios, is designed to autonomously manage debt collection by analyzing borrowers’ contexts and executing actions based on that data.
The vulnerabilities were patched in June but were publicly disclosed only in October. Independent researcher Eaton Zveare, who identified the flaws, emphasized the sensitivity of the exposed data, which includes personal financial information and extensive documentation related to unpaid debts. “Some of the largest fintech firms across India use this company to autonomously collect debts,” Zveare noted, highlighting the potential risks associated with such a data breach.
Severity of Vulnerabilities
India’s Computer Emergency Response Team (CERT-In) categorized one of the vulnerabilities as “high” severity and the other as “critical.” The high-severity flaw affected call center data, exposing hundreds of thousands of calls and the AI-generated recommendations for follow-up actions. This vulnerability could have allowed attackers to manipulate AI agents within the platform, potentially directing their own campaigns.
The critical vulnerability impacted CreditNirvana’s main platform, exposing data from three major corporate clients in the financial services sector. This breach included sensitive information such as credit reports, ID cards, case management details, and various communications related to debt collection. CERT-In warned that successful exploitation could enable attackers to impersonate users, including administrative accounts, and access or modify sensitive information.
Implications for AI and Data Security
Zveare expressed relief that the vulnerabilities were addressed quickly but noted that they may have existed in the system for years. As AI technologies become increasingly integrated into automated systems, the potential for sensitive data exposure grows. This incident serves as a critical reminder of the importance of robust security measures in AI-powered services, particularly those handling sensitive personal information.
CreditNirvana and Perfios did not respond to requests for comment regarding the incident. The rapid resolution of the vulnerabilities, however, underscores the need for ongoing vigilance in cybersecurity practices, especially in sectors dealing with sensitive financial data.
For further details on the vulnerabilities, see the full report by Wired.
As the landscape of AI-driven services continues to evolve, the implications of this breach may resonate throughout the fintech industry, prompting a reevaluation of security protocols and data management practices.


