The EU AI Regulation: A Game Changer for Insurers?

Artificial intelligence is transforming industries worldwide, and with that comes the need for regulation. This summer, the European Union (EU) introduced the AI Act, the world’s first AI legislation, to ensure the safe and responsible use of AI across all sectors, including AI in insurance.  

So, what does this mean for insurers, underwriters, and brokers in the EU? Many experts view the regulation as a crucial opportunity for insurers to lead with transparency, responsibility, and innovation.  

“The EU AI Act places a particular focus on areas in which sensitive personal data is processed. This includes life and health insurance in particular,” says Gunter Lescher, PwC Partner for Forensic Services, in an interview about the impact of the world’s first AI legislation.

Let’s break down the specific impact on the insurance industry.  

Understanding the AI Act: What’s at Stake for Insurers 

The AI Act, officially adopted on June 13, 2024 and published in July, categorizes AI systems into four risk levels: 

• Unacceptable risk 

• High risk 

• Limited risk 

• No/low risk 

Each category comes with its own set of compliance requirements. Insurers need to understand these classifications to ensure their AI applications meet the legal standards and avoid penalties. Here’s a closer look at what each risk level means and how it affects the insurance industry. 

Unacceptable Risk: Off the Table 

AI systems classified as “unacceptable risk” are banned outright. These include systems that could harm fundamental rights or manipulate behavior, such as AI-based profiling that uses sensitive data like ethnicity or social status to predict criminal activity. In the insurance world, this would cover any AI in underwriting that assesses clients based on such personal factors, particularly in life or health insurance. Automated scoring systems that rely on these variables to make decisions would also fall into this category and are strictly prohibited. 

High Risk: A Strictly Regulated Space 

AI systems deemed “high risk” are subject to stringent oversight, particularly in areas involving safety, health, and fundamental rights. While insurers may not seem directly affected by high-risk AI, systems for pricing, risk assessment, and claims management could be included. These systems must meet extensive compliance requirements, including external audits and registration in the EU’s AI database. Insurers will need to demonstrate that they handle customer data responsibly and that their AI in insurance does not infringe on fundamental rights. 

Limited Risk: Transparency is Key 

AI systems with “limited risk” are generally those that interact with customers directly, like AI-driven chatbots or systems that personalize insurance offers. For insurers using these applications, the focus is on transparency. Customers must be informed when they are interacting with AI, and any AI-generated content must be clearly labelled. While the compliance burden is lighter than for high-risk systems, insurers still need to ensure that their AI systems are transparent, and that customer data is processed responsibly. 

No/Low Risk: Minimal Oversight 

Finally, “no/low risk” AI systems — like spam filters or predictive maintenance tools — face little to no regulatory oversight. While these applications aren’t subject to specific requirements under the AI Act, companies are encouraged to adopt best practices to ensure ethical and transparent use of AI in underwriting and other insurance applications. 

Entry into Force and Application of the AI Regulation 

Although the AI Act has officially entered into force, its application will take place gradually, depending on the nature of the AI system and its risk classification. The timeline is as follows: 

• From February 2, 2025: AI systems with “unacceptable risk” are prohibited. 

• From August 2, 2025: General Purpose AI (GPAI) becomes regulated. 

• From August 2, 2026: High-risk AI systems described in Annex III must comply with regulations. 

• From August 2, 2027: High-risk AI systems listed in Annex I come under the regulation. 

This phased approach gives insurers time to assess their AI systems and ensure compliance. 

The Opportunity in AI Regulation for Insurers 

While navigating these new requirements may seem complex, there’s a silver lining. The German Insurance Association (GDV) views the AI regulation as a crucial opportunity. As Jörg Asmussen, Managing Director of GDV, highlights in Versicherungsbote magazine: “AI is also the next big development boost for the insurance sector.” The regulation could position Europe as a leader in AI governance, offering legal certainty that promotes innovation while ensuring customer trust. 

In the same magazine, Gunbritt Kammerer-Gahlan, a specialist lawyer for insurance law at Taylor Wessing reports: “Europe can become the first continent to create a high degree of legal certainty for AI applications, providing companies with planning security that they don’t have in an unregulated market.”

A New Era of AI Governance in Insurance 

The EU AI Regulation marks a significant shift in how AI is governed in Europe. Insurers, in particular, must adapt to this new landscape, ensuring that their AI systems not only comply with the law, but also build trust with customers. With AI playing an increasingly important role in everything from AI in underwriting to customer service, getting it right is crucial. 

How Sapiens Supports AI Innovation 

At Sapiens, we are committed to helping insurers harness the power of AI while navigating these new regulations. Sapiens Insurance Platform integrates advanced AI capabilities, enabling everything from personalized customer offers to automated decision making, predictive analytics, and natural language report generation. Through our partnership with Microsoft Azure OpenAI, we ensure that our solutions meet the highest standards of transparency, compliance, and performance. (For more information on our partnership with Microsoft Azure OpenAI, listen to our recent Sapiens Insurance 360 podcast, “AI Transformation: Unveiling its Generative Power with Microsoft and OpenAI,” here.)