Why AI Needs a Calculator: Building Trust in Insurance Pricing

AI has the potential to transform insurance pricing and underwriting. But should insurers trust AI to calculate the price?

According to Dani Katz, Founding Director of Optalitix, the answer is no.

In the third article of our four-part series, What Every CEO Should Know About AI, Dani explores why AI should manage workflows, interpret information and support decision-making, while trusted, deterministic pricing models remain responsible for calculations.

The principle is simple:

Let AI manage the work. Let trusted models do the maths.

Why AI shouldn't calculate insurance prices

Large language models are powerful tools for reasoning, communication and managing complex workflows. However, they are less reliable when performing complex numerical calculations.

In insurance, where pricing accuracy, transparency and governance are essential, calculations must be consistent, explainable and repeatable.

By connecting AI to established pricing engines, insurers can reduce the risk of inaccurate outputs while improving efficiency and controlling costs.

Excel still plays a critical role in insurance pricing

Despite significant investment in pricing modernisation, Excel remains central to insurance operations.

Research conducted by Optalitix and Insurance Post in 2026 found that:

• 87% of insurers surveyed use Excel for pricing
• 65% describe their Excel usage as extensive
• 76% ranked Excel highly for collaboration with underwriters
• 68% valued Excel for its transparent model logic

Rather than replacing familiar pricing tools, insurers can integrate existing Excel and Python models into scalable technology platforms.

Solutions such as Optalitix Models enable organisations to connect trusted pricing logic with modern systems, providing the foundation for AI-enabled workflows without sacrificing transparency or control.

AI as a trusted underwriting co-pilot

When AI works alongside reliable pricing models, new opportunities emerge.

AI can help underwriters explore scenarios, interpret complex results, review assumptions and identify unusual outcomes.

It can also support more consistent quality checks, flag potential pricing errors and reduce time spent on repetitive tasks.

This allows underwriting and pricing teams to focus on what matters most: applying their expertise, assessing complex risks and making better-informed decisions.

Building trust through governance

For AI to deliver meaningful value in insurance, trust must be built into the technology.

Insurers need to ensure that pricing calculations are transparent, consistent and supported by governed, version-controlled models.

As Dani explains, AI should not replace the calculations that insurers already trust. Instead, it should make those models more accessible, efficient and valuable.

The real opportunity for AI is not to replace human judgement, but to give underwriters better information and more time to apply it.

Discover why AI needs a calculator

In Part 3 of What Every CEO Should Know About AI, Dani Katz explores the architecture, governance and technology needed to make AI a reliable part of insurance pricing.

Discover how insurers can combine AI with existing pricing models to improve accuracy, reduce costs and build confidence in AI-driven underwriting.

Download the full article: Why AI Needs a Calculator

Coming next in the series: Part 4 – Architecture Is the AI Strategy, exploring why connecting systems, models and underwriting data is essential to successful AI adoption.

‍