As AI becomes embedded in core insurance and financial decision-making, explainability and accountability are emerging as critical foundations for trust, compliance and scalable innovation. Alex Johnson, Head of Insurance Solutions at Quantexa, explores why decision-level transparency and stronger data foundations are essential to turning AI from experimentation into enterprise-grade capability.
1. AI is increasingly used across insurance and financial services. Why is explainability now such a critical issue?
AI is no longer just a pilot project. It is actively driving decisions in regulatory compliance, product development, customer personalisation, claims segmentation, fraud detection and underwriting. However, as these tools move into full production, stakeholders – including leadership, regulators, customers and suppliers – demand auditability and clarity.
High accuracy alone is insufficient. It must be paired with transparency in decision chains and agentic systems. Currently, the insurance industry leverages only 10–20% of its data, often ignoring the remainder due to perceived complexity. This creates a dangerous lack of traceability. As AI impacts coverage, pricing and liability, organisations have a responsibility to show exactly how a data source was referenced and defended with evidence. Explainability across the entire data estate builds the confidence that leads to trust, allowing AI to scale from a technical tool into a core decision-making capability.
2. Why are explainability and accountability critical when using AI in these industries?
The insurance and financial sectors operate in highly regulated environments where accountability has long applied to human judgement. As AI takes on a greater role in decision augmentation, expectations have extended to automated and assisted decisions, particularly where outcomes influence pricing, financial exposure, coverage or access to services.
Regulators expect firms to provide evidence of alignment between decisions, policy and risk appetite. Auditors require traceability from source data to outcome, while customers expect fair treatment. Explainability and accountability meet these expectations while supporting responsible innovation and clearer collaboration between technology, risk, compliance and business teams.
3. How is regulatory scrutiny reshaping expectations for AI-driven decisions?
Regulatory scrutiny increasingly focuses on the transparency and fairness of decision processes rather than isolated model performance. AI-enabled decisions must be deployed across fragmented core systems and designed to rely on trusted, connected data to remain consistent over time. This includes clarity around data lineage back to a customer, supplier or third party. Controls over model inputs and production outputs must be traceable to the people and businesses affected by a decision.
There is growing emphasis on review processes conducted both during decision-making and retrospectively. Organisations that can explain individual claims, pricing and risk-based decisions not only demonstrate stronger compliance but also improve operational performance.
4. Why is decision-level transparency replacing aggregate model performance as the new standard?
Aggregate metrics, such as model accuracy, offer a high-level view of performance but provide limited insight into individual outcomes. In regulated industries, decisions occur at the individual level – for example, how a specific risk is priced or whether a particular claim is paid. Each decision carries its own implications.
Decision-level transparency focuses on understanding what drove a specific outcome, including which data points mattered, how relationships influenced the assessment and how context shaped the decision. This enables teams to see reasoning rather than scores alone, supporting clearer documentation, stronger accountability and greater trust in AI-enabled decisions.
5. How does Decision Intelligence support this shift in practice?
Decision Intelligence (DI) unifies data, entities and context into a single framework, moving beyond isolated records to reflect how real-world behaviours and interactions manifest. This creates a personalised knowledge base that feeds AI models and agentic systems, ensuring risk-based decisions are grounded in reality.
Commercial underwriting: DI automates visibility into complex corporate structures, directorships, business relationships and connected exposures. By providing underwriters and models with the same rich context, organisations can justify risk changes more easily, improve product selection and increase underwriting speed by up to 75%.
Claims assessment: Rather than analysing a claim in isolation, DI evaluates the entire network – including claimants, witnesses and suppliers. This holistic view has led to 3% improvements in loss ratios and indemnity spend, facilitating reserve releases and helping to minimise premium rate increases.
By creating a trusted feedback loop from raw data to final decision, DI surfaces data quality issues and reduces fragmentation, making every outcome easier to explain, justify and review.
6. What does an AI-ready data foundation look like in this context?
Organisations often face challenges due to fragmented systems and siloed processes. Addressing these foundations strengthens the lineage between data and decisions, improving explainability and confidence.
An AI-ready foundation rests on four pillars: trust, control, connectedness and context. Together, these ensure that every decision can be traced back to its data sources and understood within the context in which it was made.
Trust means data provenance, lineage and quality are understood and monitored. Control ensures access to data is governed, logged and policy-driven across internal teams and third parties. Connectedness allows data to be unified across people, organisations, locations and physical and digital assets, enabling models to learn from real-world relationships. Context enriches internal and external data to deduplicate customer, claimant and supplier views by up to 60%, reflecting how insurance and financial risk operate in practice.
7. How does this help insurers and financial institutions move forward with AI adoption?
Stronger data foundations and decision transparency enable insurance and banking organisations to use AI with greater internal confidence while engaging constructively with regulators through audit processes. This supports clearer governance, better risk assessment and faster learning from outcomes.
As the market matures, DI supports a transition from cautious adoption to evidence-based, confident production systems in mission-critical areas. It allows organisations to innovate responsibly and meet rising expectations for transparency. Over time, a DI approach enables organisations to convert data uncertainty into operational and competitive advantage in months rather than years.
