Actuview, the international streaming platform for actuaries, recently held a roundtable with industry insiders to discuss artificial intelligence (AI), specifically how actuaries and the insurance industry can benefit from its adoption. RGA’s Jeff Heaton served as the moderator and RGA’s Fontaine Chan was one of the panelists. The following is a transcript of her commentary.
Fontaine: While AI has been growing exponentially, the pace of the growth seems to have slowed because of resource constraints, regulatory and ethical considerations, and other technical challenges. I still see AI continuing to advance and penetrate deeper into all industries in the next five years. Companies, developers, and regulators are going to have a much stronger focus on AI governance frameworks to ensure safe, fair, and responsible use of the technology.
Fontaine: For AI implementation, it is extremely costly to train or fine tune AI models, not to mention all the resources required afterward for continuously updating, maintaining, and governing the AI system. AI models are also very complex, and it is sometimes difficult to understand the model results.
First, I would encourage companies to cultivate a culture of innovation and continuous learning. Collaboration across all the different areas is important as well, so everyone can learn from everyone, reducing the resistance to change.
There are techniques, such as explainable AI, companies can use to better understand model results. Most importantly, companies need to maintain complete and up-to-date documentation from model development, testing, and controls.
Just keep trying and I am certain companies can be successful adopting AI.
Fontaine: I can see that additional roles will be required to enhance risk management of AI models and ensure all models comply with new regulations and ethical guidelines. For new students, I encourage them to continue to be open-minded, to learn different and new technology, and keep track of all the upcoming regulations.
I can also see AI continuing to transform the actuarial field by replacing repetitive, mundane tasks with roles to interpret AI-generated insights, identify new use cases, and develop new models. RGA has started using different AI tools to help summarize documents, gather data, and other use cases.
Fontaine: Actuaries should continuously monitor data biases and data drift, and make sure to have complete documentation so all users can understand the model. Model owners should have a clear understanding of the model output and how it is developed. I think these are all essentials, apart from basing practices on all the regulations we already have, such as the EU AI Act, US-based standards, and others
The evolution of AI highlights why these ongoing, open discussions among insurance professionals are essential. Through collaboration, the industry can successfully address impending challenges and seize the opportunities of this transformative technology. Only by working together and alongside regulators can the ideal path be forged for those who create and/or use AI applications, resulting in beneficial outcomes for all, most importantly all those the industry serves.
Watch a brief clip of Fontaine’s comments below. You can see the full roundtable discussion via actuview.
Want to learn more about RGA's AI expertise and how our insights can benefit your organization? Let's talk.
