Executive summary
Insurance CFOs are under pressure to adopt AI while protecting the accuracy and integrity of financial reporting. Finance workflows with strong controls, reliable data and clear ownership provide the best starting point for AI. A workflow-first approach helps finance teams realize value now while building a strong foundation for broader AI adoption. Our four-question AI finance workflow readiness test helps CFOs decide where to deploy AI first.
AI readiness depends more on controls than the tech stack
Insurance companies are rapidly scaling AI across functions, and the finance department is becoming a key priority. In Grant Thornton’s 2026 AI Impact Survey, 38% of insurance leaders identified finance and accounting as the top function that would benefit from additional focus to meet their organization’s AI goals. This shows that insurance CFOs value AI as a way not only to reduce costs, but also to strengthen controls, accuracy and trust in financial reporting.
To keep pace with faster-moving, AI-enabled front-office operations driving significant growth, CFOs are under pressure to bring those same AI-enabled benefits to finance: accelerating the close, improving forecasting, increasing productivity and lowering the cost to operate. But deploying AI at scale isn’t their biggest challenge. It’s ensuring that AI-assisted financial reporting disclosures and regulatory filings remain accurate, complete and supported by verifiable evidence. For public companies, a single erroneous footnote or disclosure can undermine investor confidence and attract regulatory scrutiny.
As a result, CFOs must balance the promise of AI-driven efficiency with the need for transparency, auditability and controls that can withstand review by internal audit, external auditors, state departments of insurance and, for public companies, the SEC.
Before scaling AI across finance, insurance CFOs should consider this fundamental question: Which finance workflows are ready for AI?
“Readiness depends less on the technology platform and more on having strong controls, trusted data, clear ownership and an AI governance framework aligned with industry and regulatory standards,” said Thanga Ramalingam, Grant Thornton Transformation Partner. “Because traditional model risk management approaches were not designed for generative or agentic AI, insurers should establish dedicated AI governance practices that work alongside their existing risk management frameworks.”
Data and ownership don’t have to be solved everywhere first
Insurance CFOs don’t need every finance workflow to be AI-ready before they begin adopting AI. By identifying the workflows with mature financial controls, reliable data and clear ownership, they can deploy AI where it can deliver value today.
“Many leaders think AI readiness in the finance function depends on their tech stack,” said Mathew Tierney, Head of the Insurance Industry at Grant Thornton. “The stack matters, but what’s more important is whether workflow-specific financial controls are mature enough to govern AI reliably.”
Stronger AI controls are a priority for insurance leaders: The survey found the top goal of insurance finance leaders is strengthening financial controls.
But for many insurers, fragmented data across finance, actuarial, claims and policy systems, combined with inconsistent governance and unclear ownership, makes it difficult to move AI from controlled pilots to trusted, repeatable use in core finance workflows. That difficulty makes the instinct to wait for a modern tech stack first understandable, and modern finance platforms reinforce it: many of them have built-in AI capabilities, and it becomes easy to see safe AI as something that comes only after the stack is fully modernized.
But a modern platform with AI built in is not the same as governed AI. CFOs should treat any embedded vendor AI model in policy admin, claims, billing or ERP platforms as subject to the same lineage, explainability and human-override discipline as a system they have chosen and deployed themselves.
Poor data quality is an obvious objection to deploying AI without a modern tech stack, and for many organizations, it is a real constraint. Modern platforms address it at scale by consolidating fragmented sources and adding automated validation, monitoring and data lineage.
But the quality AI depends on in any single workflow comes from the same disciplines that make that workflow trustworthy in the first place: validation that catches bad inputs, clear ownership of exceptions and evidence that the data reconciles cycle after cycle. Those are workflow-level controls, not platform-level prerequisites, which is why finance can build the data quality it needs one workflow at a time.
When AI is incorporated into workflows, unclear ownership is often one of the first challenges to surface. When a model flags a transaction or proposes a match, someone still must validate that output before it reaches the numbers. That person is too often assumed rather than named.
“That's where the risk lives: an AI-influenced figure moving into the financials with no one accountable for whether it's right,” said Greg Biles, Grant Thornton Finance Partner. “A model can't own an outcome. A person has to.”
How we can help you
INDUSTRY
SERVICES
Where AI excels in insurance finance
Whether AI should support a finance workflow depends partly on the nature of the work. AI is a good fit for higher-volume, pattern-based work such as detecting anomalies, drafting reconciliations and surfacing exceptions, where fixed rules fall short. AI is a weak fit where human judgment is required or precedent is thin, such as signing off on reserve adequacy, setting IBNR reserves, or making disclosure and materiality calls. Strong-fit areas for AI in insurance finance include:
- Premium-to-cash reconciliation
- Commission reconciliation
- Statutory reporting preparation
- Reinsurance recoverable reconciliation and cedent reporting (Schedule F)
- Investment accounting reconciliation (custodian-to-GL)
- Producer and agent compensation anomaly detection
- Intercompany elimination review
- Flux and variance commentary drafting
- Close orchestration
- SOX/MAR controls evidence collection
"In our experience, finance organizations often see the greatest initial benefit in workflows with reliable data, mature controls and clear ownership,” Ramalingam said. “Reconciliation, reporting support and exception management are some areas where AI can shorten cycle times, strengthen auditability and give teams more capacity for analysis while retaining human judgment."
Deploying AI reliably depends on the strength of the financial controls that govern a finance workflow. If an organization’s finance workflows have very few mature controls, it can feel overwhelming to prepare the finance function for AI. But finance can begin deploying AI in select workflows before broader modernization is complete, as long as the targeted workflow has sufficient controls, reliable data quality and clear governance in place. Each AI-enabled workflow that delivers real efficiency and dependable results builds confidence to apply it to the next viable workflow.
Applying AI to vetted workflows can support initial adoption. But broader use cases require greater investment in a modernized enterprise data architecture with integration and governance that ensure consistency and validation across policy, claims, billing, reinsurance and investment systems. Early wins in select workflows build the case for exactly that.
Four-question AI finance workflow readiness test
Insurance CFOs should assess each finance workflow, starting with the high-volume ones, against four questions to determine where to deploy AI.
- Data traceability: Can finance trace an activity such as claims, premium, commission, reinsurance, reserves or investments from the source transaction through to the reporting outcome? Traceability is what allows finance to explain and stand behind an AI-assisted result.
- Explainability: Can finance understand, validate and document the basis for AI-generated outputs sufficiently for auditors, regulators and management? For generative and agentic AI specifically, explainability includes prompt/version control, retrieval sources, tool-use logs and a documented human-review step, not just model documentation.
- Accountability: Who owns the finance workflow, and who validates AI-assisted decisions when exceptions arise? Clear ownership keeps a person accountable within the finance operating model.
- Financial control maturity: Is the finance workflow stable and repeatable? A mature workflow shows reliable reconciliation, approval, exception management and auditability across reporting cycles.
“The finance teams getting real value from AI are the ones who start where their controls are strongest and let each workflow they get right make the case for the next,” said Anthony Lee, Grant Thornton Finance Partner.
No single workflow transforms the finance function. But as governed workflows accumulate, the effect compounds: closes that finish in days rather than weeks, capital allocation decisions informed by current, not historical, numbers, and adverse trends identified early enough for the business to change course. Each of these benefits traces back to the same foundation: workflows where controls are strong, activity is traceable from source to reporting outcome, outputs are explainable and accountability is clear.
This is where AI elevates finance’s contribution. Deployed on that foundation, AI doesn’t undercut finance's authority over the numbers. It strengthens the insight finance brings to the numbers, providing valuable inputs to decisions that shape the business.
Contacts:
Global Insurance Practice Leader
Grant Thornton International
Head of Insurance Industry, Grant Thornton Advisors LLC
Partner, Transformation
Grant Thornton Advisors LLC
Thanga Ramalingam is a Partner in Grant Thornton’s Transformation Advisory practice, specializing in AI-enabled transformation across Financial Services, including banking, asset management, private equity, and insurance.
Edison, New Jersey
Partner, Finance
Grant Thornton Advisors LLC
Finance transformation leader in insurance, Greg drives strategic growth, operational excellence, and risk-aware innovation.
Content disclaimer
This Grant Thornton Advisors LLC content provides information and comments on current issues and developments. It is not a comprehensive analysis of the subject matter covered. It is not, and should not be construed as, accounting, legal, tax, or professional advice provided by Grant Thornton Advisors LLC. All relevant facts and circumstances, including the pertinent authoritative literature, need to be considered to arrive at conclusions that comply with matters addressed in this content.
Grant Thornton Advisors LLC and its subsidiary entities are not licensed CPA firms.
For additional information on topics covered in this content, contact a Grant Thornton Advisors LLC professional.
Share with your network
Share