AI Act: The Insurance Fallout-Use-Cases & Exemplary Implementations

Not many legislations are as controversial as now-adopted EU Artificial Intelligence Act, that will regulate the usage of Machine Learning and Deep Learning methods within the European Union’s economic area.While it can be argued that usage of AI methods and processes has been slow in the insurance business, current wording suggests that also many common practices not previously considered “AI” might be subject of regulation as well. Although not all details are clear – andthat’s part of the problem - there are some areas where there will be consequences as well as an urgent need for action with respect to the actuarial practice.This follow-up training aims to provide use-cases and exemplary implementation approaches as well as impact assessments for practitioners. Participants are expected to have knowledge about Actuarial Basics and some idea about the usage and potential of Machine Learning / Artificial Intelligence in the insurance context. Understanding of the provisions of the AI Act is helpful. Laying out use cases and affected insurance processes (note that these may not even look like “AI”…), the training is structured as follows:

1. Super-short recap: Provisions of the AI act relevant for insurance
2. Conceptual toy examples
3. Impact assessment
4. Real-world use cases
5. Outlook and discussion