How AI Is Transforming Insurance Accounting Two Years After Implementation
Prepared by Hafsa Research and Analysis Company
The Core Shift: From Implementation to Optimisation
IFRS 17 became effective on 1 January 2023, replacing IFRS 4 and fundamentally reshaping insurance accounting . The standard requires entities to measure insurance contracts using current estimates of future cash flows, recognise profit over the period services are provided, and present insurance service results separately from insurance finance income or expenses.
Two years into application, the industry has entered a period of relative stabilisation. The IFRS 17 Benchmark Study 2025 by Forvis Mazars, based on 2024 financial statements from 23 major insurers and reinsurers, found growing consistency in the presentation of IFRS 17 metrics . However, significant divergence remains in key judgement areas—including discount rates, risk adjustments, and contractual service margin (CSM) methodologies—continuing to challenge comparability.
The IASB’s post-implementation review is collecting feedback globally, with potential guidance or refinements expected in 2025 or 2026 . Meanwhile, insurers are shifting focus from large-scale transformation to optimising models, systems, and disclosures.
Solution 1: Treat IFRS 17 as a Strategic Discipline, Not a Compliance Exercise
EY’s 2026 analysis of life and health reinsurance is direct: “Reinsurers that treat IFRS 17 as a strategic discipline rather than a compliance exercise can differentiate themselves” .
The CSM provides visibility into future profits, but it is highly sensitive to changes in assumptions. Amortisation patterns often diverge from actual cash flows, particularly in lumpy new business. Unlike US GAAP, which marks assets but not liabilities, IFRS 17 marks both sides of the balance sheet, amplifying swings but delivering clearer economic signals .
The practical implication: IFRS 17 is not merely a reporting standard. It is a management tool for portfolio steering, capital allocation, and investor communication.
Action step: Review your CSM roll-forward and disaggregation disclosures. Identify the assumptions driving volatility—mortality, lapse, expense, discount rates. Build a management dashboard that links CSM movements to business decisions, not just reporting outputs.
Solution 2: Address Data Fragmentation Before Scaling AI
The discipline imposed by IFRS 17 has exposed a structural challenge: life and health reinsurance relies on vast volumes of unstructured, fragmented data. Contracts, medical reports, claims narratives, and correspondence often sit outside analytical workflows, limiting insight and slowing decisions .
EY Turkey’s actuarial advisory lead describes the challenge directly: “Many companies are struggling to meet IFRS 17’s dynamic requirements using traditional, manual-heavy, Excel-based, or parametric modelling infrastructures. Producing data at the granularity required, updating projections at policy-group level, and calculating expected cash flows on a regular basis creates high workload, error risk, and operational cost pressure” .
The solution is not more manual effort—it is structural data architecture. A next-generation cash flow production tool requires: intelligent data management integrated with policy administration systems; AI-based data cleaning to detect inconsistencies and outliers; and machine learning-driven assumption generation for mortality, lapse, and expenses .
Action step: Audit your IFRS 17 data pipeline. Can you trace every input from source system to disclosure? If not, prioritise data remediation before investing in AI tools. AI amplifies existing weaknesses—it does not resolve them.
Solution 3: Build Governed AI for Auditability and Explainability
Given regulatory scrutiny, AI in insurance must be built for audit from day one. EY’s guidance is unambiguous: “Clear role definitions, explainability, monitoring, logging and alignment with model risk governance are essential. A governed reasoning layer that links signals to policy-consistent actions transforms AI from prediction to explainable, causal guidance” .
AI can convert unstructured information into pricing intelligence, surface early warning signals, and make complex models more accessible. A practical starting point is a document-intelligence backbone that normalises historical treaties and claims, supported by a governed knowledge base and retrieval-augmented generation. This allows underwriters and portfolio managers to query their own corpus and obtain citations rather than free-text outputs, while improving auditability .
Action step: For every AI use case affecting IFRS 17 outputs—CSM calculation, risk adjustment, cash flow projection—define the human checkpoint and documentation standard. If the AI cannot explain how it arrived at a conclusion, it is not ready for production.
Solution 4: Integrate IFRS 17 with Solvency II for Efficiency
Three years after implementation, the focus has shifted from compliance to integration. EY Portugal’s analysis is blunt: “IFRS 17 and Solvency II coexist, but speak different languages: Contractual Service Margin on one side, own funds on the other. This duality requires constant reconciliations, prolongs close timelines, and increases pressure on finance and actuarial teams” .
EIOPA and ASF reports confirm: transparency has improved, but operational costs remain high. Many insurers still rely on manual processes to consolidate data between frameworks. This limits agility and increases error risk .
The solution lies in automation and AI. Integrated dashboards that reconcile financial and prudential metrics are beginning to appear in European groups, offering a single view for management. This integration is not merely technical—it is strategic, ensuring solid governance and responsiveness to supervisors and investors .
Action step: Map the data overlap between IFRS 17 and Solvency II reporting. Identify where reconciliation is manual. Prioritise automation of the highest-volume, highest-error-risk reconciliation points.
Executive Checklist: IFRS 17 Optimisation Readiness
Strategic Discipline:
- □ Build CSM management dashboard linking movements to business decisions
- □ Identify key assumptions driving volatility and sensitivity
Data Architecture:
- □ Audit IFRS 17 data pipeline for traceability
- □ Remediate data quality issues before scaling AI
AI Governance:
- □ Define human checkpoints for all AI-assisted IFRS 17 calculations
- □ Implement documentation standards for AI outputs
Integration:
- □ Map IFRS 17 and Solvency II data overlap
- □ Automate high-volume reconciliation points
Closing Thought
IFRS 17 was never just an accounting standard. It was a management information revolution—forcing insurers to understand their profitability drivers with unprecedented granularity.
Two years in, the challenge has shifted. The question is no longer whether you can comply. It is whether you can extract strategic value from the data IFRS 17 generates.
As EY concludes: “The future of the insurance sector will be defined by the ability to integrate standards, optimise processes, and respond quickly to a constantly changing regulatory environment. Three years after IFRS 17, it is no longer about implementing—it is about simplifying, automating, and ensuring the right information reaches the right people at the right time” .
The insurers that master this will not just report differently. They will operate differently—and compete differently.
Prepared by Hafsa Research and Analysis Company
