Technology news and discussion on TalkFuse.
An AI Debt Collection Service Exposed 6.6 Million Accounts and 315,000 Call Recordings
Security flaws in CreditNirvana, an AI debt collection service used by major Indian fintech firms, exposed millions of borrowers’ sensitive financial records and hundreds of thousands of call recordings. The case highlights the risks of putting extensive personal and legal information into automated collection systems.
The exposed data included 6.6 million accounts, 315,000 call recordings, credit reports, identity documents, legal records, and communications such as emails, texts and transcripts. One flaw could also have allowed access as other users, including administrators, while another exposed call-centre data and could have enabled attackers to direct AI agents. Researcher Eaton Zveare worked with India’s CERT-In to address the flaws; they were patched in June and publicly disclosed in October.
In short
- CreditNirvana flaws exposed millions of borrowers’ records.
- The data included 315,000 call recordings.
- The vulnerabilities were patched in June.
Explained simply
CreditNirvana is an Indian service that uses artificial intelligence to help lenders and fintech firms manage debt collection. It can handle information about borrowers and their accounts, and support calls between borrowers and collection staff.
Debt collection can involve sensitive details, including credit histories, identity documents and legal records. Services that store and process this information therefore hold data about people’s finances as well as records of their communications with lenders.
The case raises questions about how firms protect information when they use automated systems for debt collection. It also shows why access controls matter: a flaw in such a service could expose records or allow someone to act through another user’s account.