How internal audit assesses insurtech vendor risk
Insurers increasingly depend on technology providers for claims automation, customer portals, artificial intelligence, identity verification, cloud infrastructure and data analytics. These arrangements can improve speed and service, yet they also create dependencies that may be difficult to see from a traditional procurement review. A disruption at a software provider can affect policy administration, payments, regulatory reporting and customer trust at the same time.
Internal audit has an important role in examining those dependencies objectively. Its work is broader than checking whether a vendor signed the right contract. It tests whether the insurer understands the service, data, control environment and failure scenarios associated with an insurtech relationship, then gives executives evidence for practical decisions.
Why insurtech relationships require audit attention
Insurtech vendors often provide specialised capabilities through application programming interfaces, software-as-a-service platforms and shared cloud environments. The insurer may rely on several subcontractors without having a direct relationship with each one. A vendor that appears to offer a narrow service, such as automated claims triage, may still process sensitive personal information, influence reserving inputs or connect with core policy systems.
This creates a risk profile that changes faster than the annual audit cycle. A start-up may alter its ownership, hosting arrangement, machine-learning model or financial strategy within months. A large technology provider may introduce a new fourth party or change its service architecture with limited operational notice. Internal audit should therefore assess the relationship throughout its lifecycle, rather than treating due diligence as a one-time onboarding task.
Australian insurers must also consider local regulatory expectations. APRA’s CPS 230 on operational risk management places greater emphasis on service provider oversight and critical operations, while CPS 234 remains relevant to information security capability and assurance. Privacy obligations under the Privacy Act and the Notifiable Data Breaches scheme add further importance to data handling, incident response and access management.
Establishing a risk-based assessment framework
A useful audit begins by classifying the service according to its impact on the insurer. A vendor supporting a public website may require a different level of scrutiny from a provider hosting customer records, calculating premiums or supplying a claims decision engine. Internal audit should examine criticality, substitutability, data sensitivity, transaction volume, geographic concentration and the consequences of an outage.
The assessment should include the vendor’s business model and financial resilience. A promising insurtech may depend on a small number of investors, one cloud region or a handful of key engineers. Financial distress, acquisition or loss of specialist staff could weaken service delivery even when technical controls appear strong. Audit teams can review financial statements, funding information, insurance coverage, litigation, regulatory history and evidence of business continuity testing.
Control testing should cover both the supplier and the insurer. Vendor certifications such as SOC 2 or ISO 27001 can provide useful evidence, but they do not replace a review of scope, exceptions, test dates and complementary customer controls. Internal audit should confirm that management has acted on identified gaps, obtained appropriate contractual commitments and assigned clear ownership for ongoing monitoring. Guidance on insurance audit insights can also help practitioners compare assurance approaches and identify questions for specialist providers.
Reviewing data, technology and resilience controls
Data governance is central to an insurtech review. Auditors should map what information enters the platform, where it is stored, who can access it, how long it is retained and when it is securely destroyed. The review should distinguish between customer data, claims evidence, payment information, model-training data and derived insights. These categories may have different legal, commercial and ethical considerations.
Artificial intelligence introduces additional questions. Internal audit should determine whether the insurer can explain the purpose of a model, monitor its performance and challenge unusual outcomes. Testing should consider bias, data drift, false positives, human review and the process for correcting a flawed decision. A vendor’s statement that its model is proprietary should not prevent the insurer from receiving sufficient information to meet governance, customer service and regulatory obligations.
Operational resilience requires evidence rather than policy language. Auditors can inspect recovery time and recovery point objectives, test results, backup arrangements, privileged access controls, vulnerability management and incident escalation records. They should also trace a realistic scenario, such as a cloud outage affecting Melbourne and Sydney operations during a major weather event. The exercise should show how the insurer would communicate with customers, regulators, brokers and emergency teams while restoring essential services.
Climate exposure deserves attention in the Australian market, where severe storms, floods and bushfires can generate sudden claims volumes and put infrastructure under pressure. The relationship between climate governance, reporting and supplier oversight is explored in this discussion of climate disclosure requirements. An audit may need to test whether vendors supporting catastrophe modelling, claims intake or customer communications can scale when demand rises sharply.
Building a practical vendor assurance program
Internal audit is most effective when it helps management create a repeatable process across procurement, risk, information security, legal, finance and operations. The following evidence can support a consistent review:
- A current inventory of insurtech providers, services, data flows and subcontractors
- A documented criticality rating linked to service impact and recovery requirements
- Contracts covering audit rights, data use, breach notification, continuity and exit support
- Independent assurance reports with scope, exceptions and management responses
- Records of resilience exercises, access reviews, vulnerability remediation and incidents
- Financial, regulatory and ownership monitoring for strategically important suppliers
The audit report should distinguish between control design, control operation and residual risk. A vendor may have a well-written incident response plan that has never been tested, or a strong certification that excludes the precise service used by the insurer. Clear findings should identify the business consequence, responsible executive, agreed treatment and due date rather than simply listing technical weaknesses.
Management can use a tiered monitoring model to focus effort where it matters most:
- Critical providers: quarterly risk reporting, executive oversight and regular scenario testing
- High-risk providers: annual assurance reviews, targeted control testing and remediation tracking
- Moderate-risk providers: periodic evidence refreshes and monitoring of material changes
- Lower-risk providers: proportionate due diligence, contract review and exception reporting
- Changed providers: reassessment after a merger, major incident, new data use or architecture change
This approach supports proportionality without allowing a smaller vendor to escape scrutiny simply because its contract value is low. A niche provider can still create significant concentration, privacy or customer-outcome risk if it sits inside a critical process.
Turning audit findings into stronger decisions
The value of internal audit extends beyond identifying weaknesses. Its reports can improve investment decisions by showing whether a proposed vendor can meet operational needs, regulatory obligations and customer expectations over time. Audit involvement before implementation may prevent expensive remediation, unclear ownership and rushed exit planning later.
Contracts should give the insurer practical rights, not just general assurances. Important provisions may address notification timeframes, access to records, subcontractor approval, location of data, security obligations, audit cooperation, service levels, disaster recovery and secure deletion. Exit provisions are especially important where data migration is difficult or the provider uses proprietary formats. The insurer should know how it would continue operating if the relationship ended suddenly.
Customer outcomes should remain visible in the assessment. A technically secure platform can still create harm through confusing communications, inaccessible claims processes or unexplained automated decisions. This is particularly relevant when digital services are used across metropolitan and regional Australia, where connectivity, language, disability access and digital confidence vary. Research on customer advocacy drivers provides useful context for linking vendor controls with trust and service quality.
Internal audit should report themes to the board or audit and risk committee, including concentration across cloud providers, repeated control exceptions, overdue remediation and reliance on management attestations. It can also compare the insurer’s practices with peers, regulatory developments and lessons from incidents. At professional gatherings such as the IASA Conference, finance, technology, operations and assurance leaders can exchange practical methods for dealing with these shared risks.
A mature program treats vendor assurance as a continuing discipline. New platforms, changing algorithms, outsourcing decisions and climate-related events can alter exposure without changing the supplier name. By combining risk-based testing, strong evidence, effective contracts and clear accountability, internal audit helps insurers gain the benefits of innovation while retaining control over resilience, compliance and customer outcomes.
For insurance executives and audit professionals, the next step is to review the current insurtech register against critical operations, identify the providers with the greatest customer or regulatory impact, and schedule targeted assurance work. Use the findings to strengthen contracts, test recovery arrangements and give decision-makers a clear view of the risks behind every technology partnership.