Claim settlement. Subscribe to the policy. Customer service.
Every step in insurance is a decision and every decision forms trust.
From quote to claim, insurance companies operate in a constant loop of high-risk choices. These are the moments that determine whether customers feel protected or exposed.
And this pressure is only increasing. Geopolitical instability, regulatory scrutiny, climate volatility, and changing market dynamics are colliding with growing expectations for real-time, personalized experiences.
Insurance has always been complicated. What has changed is the speed and consequences of error.
The promise of artificial intelligence – and its reality check
Artificial intelligence has captured industry attention, promising faster decision-making and better outcomes. But for many insurers, this promise has not translated into value.
according to IDC’s Data Report and “Deterministic Trust” for AI ReportOnly 7% of insurers describe themselves as “transformational.” The report also finds that only 9% of insurers have improved their use of AI, balancing power and trust.
Recent signals reinforce the illusion of the value of artificial intelligence. Major financial institutions Like Goldman Sachs, Morgan Stanley and JPMorgan Chase have publicly indicated that AI has added “essentially zero” to overall economic growth in 2025.
At the same time, failures have become difficult to ignore. A lawsuit involving Cigna’s PxDx (Procedure to Diagnosis) system allegedly resulted in over 60,000 pre-approved claims being denied in one month (a rate of one claim every 1.2 seconds). On appeal, approximately 80% of those decisions were overturned (The class action lawsuit was filed just last year).
Unfortunately, “good enough” AI will not succeed in the insurance industry.
Where real value is gained and lost
The risks are huge. Bad decisions in claims alone represent Losses estimated at $170 billionwith another $160 billion tied to underwriting inefficiencies.
But the upside is quite real. Insurers that lead AI significantly outperform their peers – in some cases, achieving more than 6x total shareholder return.
Striking a balance between using AI and avoiding its downside risks will not be an easy task. Linking decisions across the insurance value chain will include alignment with strategic goals and a commitment to data governance and cultural evolution.
From theory to implementation
Although it would be easy to argue assumptions and throw around buzzwords like “customer” or “Generative artificial intelligenceLet’s explore three real-world use cases: real insurance companies, real outcomes.
DB Insurance protects capital by fighting fraud
DB Insurance protects more than 10 million customers in Korea. Like many insurance companies, bad actors found weaknesses in their operations and exploited the weaknesses to defraud the insurance company.
With so many customers and so many claims, human investigators were overwhelmed, often missing more complex patterns, and outmaneuvered by more sophisticated fraud networks.
Therefore, DB Insurance is committed to fighting with fire. They partnered with SAS to develop Korea’s first AI-powered fraud detection network, the DB T-System. By using machine learning to unify operational and informational data on a single platform, decades of policy, claims and customer information are brought together.
The AI learns with each new case it processes, improving its ability to distinguish between legitimate claims and fraudulent claims. With implementation, analysis time decreased from hours to just two minutes. Detection accuracy improved by 99% and cases treated increased 30-fold.
Investigators can now act proactively to prevent fraud. DB Insurance can now protect people and businesses, offer fairer premiums and resolve claims faster.
Neova Sigorta generates revenue through premium AI-based modeling
The Turkish Motor Property & Casualty company has seen tie-up quote rates less than half the industry standard, with no sales growth and weak profitability. With a new mandate from the CEO, they sought a single platform to increase sales, improve segmentation and combat fraud.
The company’s pricing method was working against their business goals – outdated GLM algorithms. They then began investing in machine learning-based pricing to enable smarter underwriting, agents and clients. This initiative had the potential to offer better insurance rates to 95% of its customers. Additionally, conservative forecasts indicate a sales increase of up to 15% with the insurer’s combined ratio declining by a staggering 10%.
As a result of adopting a machine learning-based approach, Neova Sigorta notes that not only has its success rate increased – which translates into increased sales – but those rates are now more competitive, with an approximately 9% reduction in overall premiums. Their AI-powered approach to underwriting and risk selection has delivered returns so compelling that they are now rolling out AI solutions across the entire company.
ERGO crowns customers by pioneering customer intelligence
ERGO is one of the largest insurance groups in the world and has been a customer of SAS for more than 35 years. They keep pushing the envelope artificial intelligence (The last one Deploy an enterprise-wide generative AI assistant to 28,000 employees). And they continue to innovate. Their quest to transform customer data from mundane CRM functions into a revolutionary, comprehensive digital strategy requires more than just artificial intelligence. Creating a true omnichannel experience means blending technological innovation with human authenticity.
ERGO’s strategy combined online and offline channels, providing fast, convenient and relevant experiences to its customers. Drawing on their leading expertise in machine learning and Deep learningtheir approach improved the next best offer, next best action, customer service and processing time across all customer engagements, regardless of channel.
This forward-looking, customer-obsessed philosophy has paid off. ERGO has achieved its strongest new customer business in a decade for two years in a row.
Decisions build or break trust
Insurance works on results.
Every claim approved or denied, every risk priced, and every interaction with customers shapes whether insurers protect capital, secure revenue, and retain customers.
The insurers profiled here aren’t experimenting with AI just for efficiency’s sake. They work to improve the quality of decisions – reducing fraud, pricing risk more accurately and delivering relevant customer experiences.
When AI lacks consistency or transparency, it puts critical insurance decisions at risk.
In a regulated industry, a decision you can’t explain is a decision you can’t defend.
Because in insurance, the question is whether the decisions it drives can be trusted.
If you’d like to learn more, consider joining us at SAS Innovate, April 27-30, or at one of the many stops on the SAS Innovate Global Tour.
If you want to keep up with the latest technological innovations such as quantum AI, agent AI, or synthetic data, you can subscribe to SAS channel on YouTube.
