Insurance Europe has voiced concerns over the European Insurance and Occupational Pensions Authority (EIOPA)´s draft opinion on artificial intelligence governance and risk management, suggesting it could unintentionally result in hidden compliance burdens for insurers. While EIOPA´s intent is to clarify how existing regulatory frameworks apply to Artificial Intelligence, Insurance Europe warns that the opinion’s language may be interpreted by supervisors as establishing de facto new requirements, thereby extending expectations beyond current legislation.
The industry group is calling for a more robust application of proportionality within the draft opinion, emphasizing the importance of differentiating between various types of Artificial Intelligence applications—such as those used internally versus those directly facing customers—and distinguishing between organisations that develop Artificial Intelligence versus those that deploy it. Insurance Europe also highlights the significance of recognizing distinctions between traditional machine learning and the evolving class of generative Artificial Intelligence systems.
Recent EIOPA data cited by Insurance Europe shows that 50% of non-life insurers and 24% of life insurers across Europe currently use Artificial Intelligence, with additional planned adoption projected in the next three years. As Artificial Intelligence becomes increasingly prevalent, clarity in regulation is seen as critical to prevent stifling innovation through unnecessary administrative burdens. Generative Artificial Intelligence tools are already yielding efficiencies in the insurance sector, including document customisation, automation in regulatory reporting, and analysis of unstructured data, though they introduce novel risk factors requiring careful oversight. Insurers, such as those in the Lloyd’s of London market, are even offering specialized coverage for losses arising from Artificial Intelligence malfunctions or unpredictable outcomes, reflecting the complexity of these risks.
Insurance Europe also flags the potential issue of dual supervision, where compliance with Artificial Intelligence governance may fall under multiple authorities, such as those enforcing the European Union’s AI Act alongside insurance sector-specific regulators. To alleviate regulatory uncertainty, it urges EIOPA to provide explicit guidance on managing these overlaps. The group further stresses the necessity for EIOPA to focus its opinion on interpreting existing frameworks, such as Solvency II, the Insurance Distribution Directive, General Data Protection Regulation, and the Digital Operational Resilience Act, which already encompass mechanisms for Artificial Intelligence governance, risk management, and consumer protection within the industry.
