Artificial intelligence is transforming actuarial practice, creating new opportunities while raising questions about professional responsibility, governance, and regulation. Two recent publications from the International Actuarial Association (IAA) and the Actuarial Association of Europe (AAE) address these developments from different perspectives, offering insights into how actuaries can navigate the increasing use of AI while maintaining professional standards and ensuring responsible model development and application.
The IAA has published a paper titled “Professional Considerations for Actuaries Interacting with Artificial Intelligence”, developed by its AI Task Force. Intended for actuaries who develop, validate, or use AI systems and models, the paper provides practical questions and considerations rather than a formal compliance checklist. It covers topics such as model development, interpretability and explainability, bias mitigation and fairness, data privacy and security, and effective communication of AI-generated insights. A central message is that actuaries remain professionally responsible for their work, including results produced with AI assistance. The publication also highlights the importance of continuous learning and collaboration with AI specialists as technology evolves.
Meanwhile, the AAE has published a note titled “The AI Act and Traditional Actuarial Models: A Fundamental Rights Perspective”, prepared by its Artificial Intelligence and Data Science Working Group. The paper examines the treatment of traditional actuarial techniques, particularly generalized linear models (GLMs) and generalized additive models (GAMs), under the European Union’s AI Act. It argues that human-supervised applications of these established statistical methods should not automatically be classified as high-risk AI systems, as existing European insurance regulation, data protection requirements, and professional standards already provide substantial safeguards for fundamental rights in insurance pricing. The working group emphasizes the need for proportionate regulation, particularly concerning data governance, complex machine learning components, and potentially discriminatory outcomes, and calls for closer cooperation between European regulators and the actuarial profession.
Together, these publications illustrate two important dimensions of the profession’s response to AI: maintaining professional accountability as new technologies become part of actuarial work and ensuring that regulatory frameworks appropriately reflect the characteristics and risks of different modeling approaches. For actuaries worldwide, both contributions offer perspectives on the evolving relationship between technological innovation, professional judgment, and responsible actuarial practice.
Explore the full publications: IAA: Professional Considerations for Actuaries Interacting with Artificial Intelligence | AAE: The AI Act and Traditional Actuarial Models
