Insurance Europe has responded to the European Insurance and Occupational Pensions Authority (EIOPA)’s draft opinion on artificial intelligence (AI) governance and risk management, cautioning that while the opinion is intended to clarify how existing rules apply to AI, it may inadvertently establish new compliance burdens for insurers.
Although the opinion explicitly states it is not designed to impose new regulatory requirements, Insurance Europe points out that the wording could be interpreted by supervisors as creating de facto obligations. The industry body warns that such interpretations may lead to expectations that go beyond the scope of current legislation.
In its submission, Insurance Europe calls for a stronger application of proportionality within the opinion. It emphasises the need to account for differences in how AI is used, such as between internal processes and customer-facing tools, and between those who develop AI systems and those who deploy them.
The organisation also highlights distinctions between types of AI, noting the importance of separating traditional machine learning applications from generative AI systems.
In the context of EIOPA’s ongoing digitalisation efforts, recent data shows that 50% of non-life insurers and 24% of life insurers in Europe are currently using AI, with a further 30% and 39%, respectively, planning to adopt such technologies within three years. This growing adoption of AI across the sector underlines the importance of regulatory clarity without overburdening firms.
The emergence of generative AI tools is creating new efficiencies for insurers, such as enabling customisation of policy documents, automation of regulatory reporting, and analysis of unstructured data sources. These developments, however, come with novel risks that require careful regulatory alignment.
The issue of reliability and potential liabilities associated with AI is also gaining prominence. In response to growing concerns over chatbot errors and unpredictable AI behaviour, insurers operating in the Lloyd’s of London market have introduced coverage specifically for losses arising from AI-related malfunctions.
Another area of concern is the potential for dual supervision. Insurance Europe notes that in some jurisdictions, oversight of AI compliance may be split between multiple authorities, such as those enforcing the EU’s AI Act and those supervising sector-specific insurance rules.
It has asked EIOPA to explicitly address how such overlaps should be handled to avoid regulatory uncertainty.
Insurance Europe has also underlined the importance of maintaining a focus on existing frameworks. It says that the opinion should concentrate on interpreting how current sectoral regulations – including Solvency II, the Insurance Distribution Directive, the General Data Protection Regulation, and the Digital Operational Resilience Act – already provide mechanisms for governance, risk management and consumer protection in relation to AI use.
What are your thoughts on this story? Please feel free to share your comments below.
