Cyber Insurance for NBFCs After an AI-Assisted Breach
C
Coverbiz Team
Author
The reported Delhi-NCR incident shows why cyber insurance for NBFCs cannot be assessed only by asking whether a company has suffered a data breach. The more difficult question is whether an attacker reached a loan-origination or disbursement system, changed an approved borrower’s payment destination, bypassed an approval control, or caused employees to authorise a transfer. A November 17, 2026 Indian Express report described police investigations into 12 NBFCs in Delhi-NCR, including one case where investigators detected traces of an AI tool. The reported ₹65 crore figure should remain attributed to that report and should not be presented as independently confirmed.
For NBFC founders, CFOs and risk teams, the immediate priorities are to establish the facts, preserve evidence, stop further disbursements where necessary, meet regulatory reporting duties and map the loss against the correct insurance section. Cyber cover, commercial crime cover and fraud extensions can overlap, but they are not interchangeable.
What the reported Delhi-NCR NBFC breach may mean for financial businesses
The reported pattern is particularly relevant to lenders because an attacker may not need to compromise every customer account. If an attacker can find loans that have already been approved but not disbursed, alter the beneficiary details or send a valid-looking command to a disbursement API, the loss can occur inside an apparently legitimate workflow.
The report described allegations that attackers used weaknesses in public-facing software, accessed approved but undisbursed loans and routed money to mule accounts. It also described the use of cloud infrastructure whose IP addresses had been allowed by the NBFC’s firewall. These are reported investigative findings, not a technical conclusion that applies to every NBFC or every incident. The underlying businesses, police records, logs, transaction trails and recoveries still need to be assessed case by case.
The regulatory direction is clearer. The RBI’s 2026 Cybersecurity, Technology: Risk, Resilience and Assurance Framework for NBFCs requires incident-management processes, audit trails, a Board-approved cybersecurity policy, a Cyber Crisis Management Plan, access controls and reporting of cyber incidents to RBI within six hours of detection on the DAKSH platform. It also specifically identifies threats such as business email fraud, phishing, identity fraud and password-related fraud.
The same RBI direction requires audit trails that can support forensic evidence and dispute resolution. For an NBFC, that means retaining more than application logs. The investigation may need API requests, source IPs, cloud-account details, identity-provider events, privileged-user activity, approval records, beneficiary changes, bank acknowledgements and reconciliation reports.
The practical lesson for nbfc cyber security is to connect technology controls with lending controls. A secure perimeter is not enough if an internal or trusted cloud path can initiate a payment without a second approval, a transaction-risk check or an independent reconciliation.
Cyber insurance for NBFCs: what the policy is designed to address
Corporate cyber insurance generally has two broad parts. First-party cover addresses the insured business’s own response and interruption costs. Third-party cyber liability cover addresses claims or proceedings brought by customers, business partners, regulators or other affected parties, subject to the wording.
Depending on the policy, possible first-party sections may include digital forensics, legal advice, crisis management, customer notification, data restoration, cyber extortion, business interruption and extra expense. Corporate cyber insurance for Indian businesses can also include network security and privacy liability, media liability or regulatory-response costs. Coverbiz’s corporate cyber insurance page describes these as possible policy features, but the actual schedule and insurer wording control the claim.
For an NBFC, the main questions are practical:
Does the policy cover forensic investigation of a compromised loan or payment platform?
Are incident-response vendors available through a panel, and must the insurer approve their appointment?
Is business interruption triggered by an outage, unauthorised access, malicious code or another defined event?
Are customer notification, call-centre, legal and public-relations costs covered?
Does cyber liability respond to privacy, network-security or regulatory allegations?
Are outsourced technology providers, lending-service providers, cloud platforms and subsidiaries included?
Do sub-limits apply to forensics, notification, restoration, business interruption or liability defence?
The Coverbiz guide on what Indian businesses should check before buying cyber insurance explains why first-party and third-party limits should be separated and why waiting periods, retroactive dates, consent provisions, vendor scope and exclusions require careful review.
Insurance does not replace NBFC controls. The RBI direction itself refers to information-security risk management frameworks that include internal controls, processes and applicable insurance covers. Insurance is therefore one part of a wider risk-management structure, not a substitute for secure APIs, effective authentication, tested recovery and payment governance.
Could diverted approved loans fall under cyber, crime or fraud cover?
This is the central coverage question in an incident involving diverted loan disbursements. The answer depends on what happened technically and operationally.
1. Cyber cover
A cyber policy may respond where the loss follows unauthorised access, malicious code, compromise of credentials, manipulation of data or another covered cyber event. It may also cover investigation, system restoration, interruption and liability even if the direct loan principal is not insured under the same clause.
The policy may, however, define financial loss narrowly. A clause for electronic theft may require fraudulent data to have been entered into the insured’s system or network. That may be different from an employee approving a payment after receiving a false instruction. The distinction is explained in Coverbiz’s guide to business cyber-fraud insurance and finance-team response.
2. Crime or fidelity cover
A commercial crime or fidelity policy may be more relevant where the loss involves theft, employee dishonesty, computer fraud, funds-transfer fraud or fraudulent instructions. The exact terminology varies. Some policies respond to direct financial loss, while others exclude losses caused by an employee’s voluntary transfer or require a specific endorsement for social engineering.
The RBI’s Fraud Risk Management Directions for NBFCs require a Board-approved fraud-risk policy covering prevention, early detection, investigation, recovery and reporting. That governance process should operate alongside insurance review. An NBFC should not wait for a coverage decision before reporting suspected fraud, contacting banks or preserving evidence.
3. Social engineering or fraudulent transfer extensions
Social engineering usually describes the method used to manipulate a person into taking an action. Payment diversion describes the result. The two may occur together, but a policy clause may treat them differently from an unauthorised technical transfer.
Ask the insurer to identify the exact clause for:
electronic theft;
computer fraud;
funds-transfer fraud;
fraudulent communications or instructions;
social engineering;
employee dishonesty or fidelity; and
loss caused by a compromised third-party service provider.
Do not assume that a broad “cyber” label covers the loan principal. Review the schedule, definitions, exclusions, single-loss limit, aggregate limit, deductible, notification deadline and proof-of-loss requirements.
The first 24 hours: preserve evidence, contain access and notify the right teams
A suspected diversion should be handled as a combined cyber, fraud and business-continuity incident.
Stop or restrict further disbursement risk. Consider a controlled pause for affected products, APIs, beneficiaries, credentials or integration paths. Do not make broad system changes before forensic advice if they may overwrite evidence.
Contact the bank immediately. Request payment holds, recalls, beneficiary-account flags and transaction tracing. Record every bank reference number.
Preserve evidence. Export relevant application, API gateway, identity, endpoint, cloud, firewall, VPN, database, approval and payment logs. Preserve timestamps in a consistent time zone and maintain a chain-of-custody record.
Create one incident chronology. Record when the event was detected, who detected it, affected systems, transactions, beneficiaries, approvers, containment actions, bank communications and recovery attempts.
Notify the insurer or broker promptly. Early notice may be a policy condition. Ask before appointing vendors, admitting liability, making settlements or incurring large response costs where prior consent is required.
Involve the right decision-makers. The incident team should include technology, information security, finance, compliance, legal, fraud risk, communications and senior management. Add the relevant lending-service provider or cloud provider under a controlled evidence request.
Assess regulatory and law-enforcement reporting. The RBI framework requires applicable NBFCs to report cyber incidents within six hours of detection. The CERT-In Directions dated April 28, 2022 require covered entities to report listed incidents within six hours of noticing them or being informed of them, and require ICT logs to be maintained securely for a rolling 180 days within India. The exact reporting route should be confirmed by the compliance and legal teams.
Report suspected financial fraud without waiting for the insurance review. The RBI’s fraud-risk FAQ says regulated entities should immediately report fraud incidents to law-enforcement agencies, subject to applicable law, and advises reporting fraud involving ₹1 lakh or more to law-enforcement agencies.
Controls NBFCs should review after an AI-assisted attack
The reported use of AI should not lead to a narrow focus on the tool. The control question is whether automation made an existing weakness easier to exploit or scale.
Access and identity
Remove dormant accounts and review privileged access.
Enforce multi-factor authentication for administrators, cloud consoles, VPNs, payment systems and developer tools.
Separate developer, operations, approval and production-disbursement privileges.
Use just-in-time or time-bound privileged access where practical.
Review service accounts, API keys, secrets and long-lived tokens.
Alert on unusual logins, new cloud accounts, impossible travel and changes to trusted IP ranges.
The RBI’s 2026 framework says access to information assets should be allowed only where a valid business need exists and that elevated system activity should be logged and periodically reviewed. This is directly relevant to cyber risk for NBFCs because trusted identities and trusted infrastructure can be abused without a conventional perimeter alert.
Loan-disbursement and payment approval
Require two-person approval for beneficiary changes and high-value disbursements.
Separate loan approval from payment release.
Add a cooling period for changes to borrower bank details.
Use transaction-risk rules for first-time beneficiaries, unusual velocity, multiple loans to one account, out-of-hours activity and unexpected IP or device changes.
Verify that the beneficiary account belongs to the approved borrower before release.
Reconcile disbursements daily, not only weekly or fortnightly.
Alert when one source, token, IP range or service account initiates an unusual number of disbursement requests.
Do not treat a whitelisted IP as proof that a request is legitimate.
Application, API and cloud controls
Test public-facing applications and APIs for authentication, authorisation and input-validation weaknesses.
Review whether an API can change a beneficiary or disbursement instruction without a fresh approval.
Replace broad IP allowlists with stronger identity, device, certificate and request-signing controls where suitable.
Confirm that vendors notify the NBFC without undue delay and support RBI and CERT-In reporting.
Test recovery from clean backups and retain logs in a form investigators can use.
Questions to ask an insurer about sub-limits, exclusions and panel responders
Before renewal or placement, ask for written answers to these questions:
Does the policy cover direct loss from computer fraud, electronic theft, fraudulent instructions or social engineering?
Is a diverted loan disbursement treated differently from a payment made after employee deception?
Are the relevant benefits inside the main aggregate or available as additional limits?
What are the single-loss limits and deductibles for funds-transfer fraud, forensics, business interruption and cyber liability?
Is there a waiting period before interruption losses are calculated?
Are known vulnerabilities, unpatched systems, weak access controls or failure to follow a payment procedure excluded?
What happens if the compromised system belongs to a lending-service provider, cloud host or other outsourced technology vendor?
Which forensic firms, breach counsel, crisis advisers and negotiators are on the panel?
Can the NBFC appoint its own responder, and does the policy require prior written consent?
What notice must be given after discovery, and what records are required to prove the loss?
How do other-insurance, subrogation and recovery provisions work if the bank, vendor or payment intermediary recovers part of the money?
Are regulatory investigations, customer claims and contractual liabilities covered, and under which section?
These questions are especially relevant to cyber fraud insurance for NBFCs because the financial loss may sit between cyber liability, crime insurance, fidelity cover and operational-fraud controls.
FAQ: Does cyber insurance cover social engineering and fraudulent transfers?
Does cyber insurance cover social engineering and fraudulent transfers?
It may, but not automatically. A policy may distinguish unauthorised system activity, electronic theft, fraudulent communications, social engineering and employee-approved transfers. Check the exact insuring clause, definitions, endorsements, sub-limits, deductible and claims conditions.
Could diverted approved loans be covered under a cyber policy?
Possibly, if the loss resulted from a covered cyber event or electronic theft clause. However, a fraudulent transfer may instead require crime, fidelity, funds-transfer fraud or social-engineering cover. The facts and issued policy wording decide the route.
What should an NBFC do in the first 24 hours after suspected compromise?
Contain affected access without destroying evidence, contact banks to stop or trace payments, notify the insurer or broker, preserve logs and approval records, involve legal and technology teams, and assess reporting to RBI, CERT-In and law-enforcement authorities.
Does RBI require NBFCs to report cyber incidents quickly?
The RBI’s 2026 cyber-resilience directions require applicable NBFCs to report cyber incidents to RBI within six hours of detection on the DAKSH platform and to proactively notify CERT-In. The applicable direction and the facts of the incident should be reviewed with the NBFC’s compliance and legal teams.
How to prepare a cyber-insurance submission after a suspected breach
An insurer will need more than a headline loss estimate. Prepare a controlled submission containing:
a short incident summary with confirmed facts separated from assumptions;
the date and method of discovery;
affected systems, APIs, vendors and data categories;
the number and value of disputed transactions;
bank references, police complaint details and regulatory notifications;
relevant logs and forensic findings;
the incident-response timeline;
recovery, recall and freeze attempts;
copies of the policy schedule, wording and endorsements;
current security controls and known gaps; and
the expected costs for forensics, legal advice, restoration, notification, interruption and liability defence.
Avoid describing the loss as “covered” before the insurer reviews the facts. Equally, avoid delaying notice while the investigation is incomplete. A careful first notice can state what is known, what remains under investigation and which policy sections may be relevant.
NBFCs should also update their next submission with the controls that matter to a lender: privileged-access reviews, API security testing, cloud-configuration monitoring, beneficiary verification, dual approval, reconciliation frequency, log retention, incident exercises and vendor oversight. This helps the insurer assess the actual exposure rather than relying on a generic technology questionnaire.
If your business needs a structured review, you can get a business insurance quote from Coverbiz for corporate cyber insurance, incident-response requirements and the distinction between cyber, crime and other relevant covers. Cover, exclusions and claim decisions depend on the insurer and the policy terms.
Does cyber insurance cover social engineering and fraudulent transfers?
It may, but not automatically. A policy may distinguish unauthorised system activity, electronic theft, fraudulent communications, social engineering and employee-approved transfers. Check the exact insuring clause, definitions, endorsements, sub-limits, deductible and claims conditions.
Could diverted approved loans be covered under a cyber policy?
Possibly, if the loss resulted from a covered cyber event or electronic theft clause. However, a fraudulent transfer may instead require crime, fidelity, funds-transfer fraud or social-engineering cover. The facts and issued policy wording decide the route.
What should an NBFC do in the first 24 hours after suspected compromise?
Contain affected access without destroying evidence, contact banks to stop or trace payments, notify the insurer or broker, preserve logs and approval records, involve legal and technology teams, and assess reporting to RBI, CERT-In and law-enforcement authorities.
Does RBI require NBFCs to report cyber incidents quickly?
The RBI’s 2026 cyber-resilience directions require applicable NBFCs to report cyber incidents to RBI within six hours of detection on the DAKSH platform and to proactively notify CERT-In. The applicable direction and the facts of the incident should be reviewed with the NBFC’s compliance and legal teams.
Do you have more questions?
Contact us for any queries related to business insurance, coverages, plans and policies. Our insurance experts will assist you.