Scaling AI with confidence: the smart approach to unlocking business value

This article is based on a discussion between Penny Seach, Chief Underwriting Officer at Zurich, and Ericson Chan, Group Chief Information and Digital Officer at Zurich, at Zurich’s Global Risk Management Summit 2025.

AI’s unmatched rise

AI has advanced at a pace unlike any other technological shift in history. In just three years, it has transformed from a tool for experimentation into an indispensable part of modern life. From automating routine tasks to generating human-like reasoning and decision-making, AI is already a powerful agent of change in professional and personal lives.

The speed of change is exponential. AI is teaching itself to improve – a self-reinforcing cycle of learning that sets it apart from all previous technological revolutions. Progress is further reinforced by a seismic level of investment: in 2024 alone, the six largest technology companies collectively spent over $200 billion (S&P Capital IQ 03/25) building AI infrastructure.

Unlocking business opportunities

With the cost of deploying AI falling dramatically in the past three years, AI’s potential to unlock business value is immense - McKinsey estimates that its impact on productivity could add trillions of dollars to the global economy annually.

The technology offers the potential to streamline operations, enhance customer experience and free up employees to focus on higher-value, strategic work. It can assist in analyzing large datasets, identifying patterns, and generating insights that humans alone would struggle to identify, enabling faster, data‑led operations. These outputs are intended to inform and support decision‑making processes, which remain subject to established controls.

The use cases are almost limitless and vary across industries. In commercial insurance, for instance, Zurich has identified underwriting, distribution, claims and resilience solutions as the most promising areas for extracting value, and it has deployed proprietary AI-powered platforms across these areas to improve customer outcomes.

For example, Zurich uses AI to help extract information from unstructured email data and feed it into submission and pricing systems to speed up response times. Meanwhile, underwriters are supported by bespoke chatbots that provide instant access to relevant documentation and guidance, enhancing both efficiency and accuracy.

The benefits go beyond efficiencies. For multinational customers with complex global programes, for example, we use AI to strengthen contract certainty by ensuring that local policy wordings are accurate, consistent and fully aligned with master terms. AI can also be used in, for example, catastrophe claims handling to automatically identify and validate claims, enabling faster and more accurate processing.

Developing maturity with confidence

Despite the endless opportunities, few established businesses have become truly AI-native. While many companies are experimenting with AI, most are stuck in pilot phases and are yet to scale up or drive true value.

Businesses are cognizant of the need to realize its potential safely and responsibly. Caution stems from three primary areas: data privacy and security; the reliability and accuracy of model outputs; and workforce adaptation challenges.

Businesses are wise to be prudent, but none of these challenges are insurmountable. With responsible data stewardship, transparent governance structures and organizational training, businesses can move responsibly from experimentation to execution.

Building AI maturity is a gradual, structured process. It requires clarity of purpose, disciplined experimentation and a strong foundation of governance and skills. Successful organizations will take a deliberate approach, underpinned by five key steps:

1. Democratize AI
   AI should not be confined to a small group of specialists. Democratizing access means equipping employees across all functions with tools, training and confidence to use AI safely and effectively. This drives innovation from the ground up and embeds AI thinking into the culture.
2. Identify high value use cases
   Rather than starting with the technology, start with the business challenge. Map the value chain and pinpoint where AI can deliver the greatest impact – whether through automation, insight generation or risk reduction. Focusing on a small number of high-value areas enables faster, measurable returns and helps build internal credibility.
3. Invest in robust data.
   As with any model, the output of AI systems is only as good as the input. Ensuring data quality is therefore essential when investing in AI. Data should be relevant, accurate, correctly labelled, up-to-date and consistent. Businesses must ensure datasets are not unintentionally biased or unfair.
4. Experiment and learn safely
   Pilot projects should be run in controlled environments with strong data governance and human oversight. Test and fail quickly, then integrate and scale the solutions that are truly adding value. Quantifying results, such as productivity gains, cost savings and customer outcomes, helps secure senior management buy-in and future investment.
5. Invest in skills and change management
   Upskilling employees is critical. Training should go beyond technical skills to include critical thinking, ethical awareness and effective prompting. Change management should be proactive, positioning AI as a tool for empowerment, not replacement.

Data security should be prioritized at all stages. At Zurich, we take this commitment extremely seriously, through a multilayered security approach and a commitment to transparency. Our approach to a safe, responsible and customer-centric use of AI technologies is based on the STAR framework: Safety, Transparency, Accountability and Reliability.

As organizations move from experimentation to scaling, they should also explore how entire business models can be reimagined using advanced, end-to-end AI systems. Agentic AI, for instance, opens opportunities to automate complex workflows, self-optimize processes, and deliver adaptive decision-making – all within a robust governance framework.

Underwriting implications

For insurers, AI represents both an opportunity and a new class of risk. AI is ubiquitous, interconnected and it is already part of the risk fabric. However, it is not yet clear how this will manifest in terms of business consequences or, therefore, how it will impact insurance outcomes.

Underwriters will increasingly need to assess an organization’s AI maturity level, addressing questions such as, how well is AI governed? What safeguards are in place to ensure responsible use?

Underwriters will be looking for companies that can demonstrate robust governance frameworks, reliable data management and strong accountability structures – as well as a demonstrable commitment to transparency.

As AI adoption accelerates, risk professionals will need to balance innovation with prudence, ensuring that enthusiasm for new tools does not outpace the controls required to manage them.

Speed, insight and efficiency

AI’s rise has sparked both excitement and uncertainty. It is reshaping industries, creating new opportunities and challenging how businesses operate. While its rapid evolution can feel unsettling, one thing is clear: the AI revolution is well underway.

For organizations willing to approach AI strategically – identifying where it adds real value, embedding strong governance and building employee confidence – the rewards are substantial. Those that delay may find themselves at a competitive disadvantage as peers harness AI to drive speed, insight and efficiency.

In insurance, AI has immense potential to remove friction from the industry, injecting speed and accuracy into a range of processes, freeing employees to focus their time where the customer values it most. At Zurich, we are passionate about realising this potential. The improvements we are already seeing for our customers and brokers are just the beginning as we seek to accelerate an even deeper integration of AI across our business.

In a world increasingly defined by intelligent systems, the smartest move businesses can make is to combine innovation with integrity – unlocking AI’s potential responsibly, sustainably and with lasting impact.

Originally published on Commercial Risk on March 24, 2026