Description: Artificial intelligence (“AI”) adoption in the insurance industry is increasing. One known risk as adoption of AI increases is the potential for unfair bias. Central to understanding where and how unfair bias may occur in AI systems is defining what unfair bias means and what constitutes fairness.

Description: Building new models or migrating from one system to another is hard! During this session, you will hear from your colleagues who have gone through such projects. You will gain a better sense of what to expect, what to avoid, and what to prioritize for such projects. The discussion will be system agnostic. Learn about pitfalls and how to avoid them!

Description: Learn how insurers provide new opportunities for efficiency and value creation from product design and underwriting through customer engagement and claims. Glimpse some industry use cases for Gen AI and its outcomes as observed by the first movers.

Description: The global doubling of information content has been estimated to occur at a rate of every 12 hours, and conceivably faster. This ‘infowhelm’ is a double-edged sword for insurers. While maximization of the global knowledge base can fuel innovation, it can occur only in the context of reasonable cost for a higher benefit. The insurance industry – underwriters and actuaries in particular – are challenged to constantly examine new innovations that promise to raise confidence and protective value for insurability, within legal limits. This report aims to be the first of several to alert the SOA of promising innovations for the profession.

Description: Today's best AIs often match human experts on routine tasks - but AI automation success requires skilled implementation. Nathan will present a field-tested, three-step method for automating cognitive work: (1) understanding what AI can do & choosing the right tasks, (2) documenting processes in the form of ""gold standard examples"", and (3) systematically optimizing AI performance – all using concrete examples from actuarial practice. Master this approach and amplify your own expertise by automating repetitive work with AI. Nathan Labenz, founder and former CEO of Waymark, is a technology entrepreneur and artificial intelligence analyst. With a background in science and an eye for innovation, Nathan led Waymark from its inception to its status as a trailblazer in generative AI-powered video creation. Currently, he hosts 'The Cognitive Revolution' podcast, where he explores the transformative impact of artificial intelligence on work, life, society, and culture. Through his conversations with notable builders, researchers, and investors, Nathan helps business, policy, and academic leaders stay up to date with AI developments and implications.

Description: The actuarial profession is at a tipping point and a combination of disruptive technologies and external pressures are creating a unique opportunity to move from results producer to insightful adviser. For a prudent but forward-looking professional, this is exciting and scary in equal measure. The ability to process large masses of business data to make more accurate calculations has become one of the most sought-after skills for leaders across all industries, including insurance.<br><br>The presenters will introduce current data technologies used to manage, process and visualize data. They will cover solutions like Microsoft BI, Tableau, Amazon Redshift, Hadoop, Alteryx, Python, etc. and their applications in data, process improvement, automation, reporting and predictive analytics. In particular, they will explore management reporting and how it leverages data analytics to help business leaders with strategic decisions making, how to evaluate the effectiveness of management reporting structures, drivers of change and ways to improve them, and proposed steps toward entrenching this dynamic and insightful approach to management reporting in which insurance leaders can have confidence.

Description: This article introduces IT DevOps to actuaries and how it is applicable to them to improve their modeling efforts and make them more efficient.

Description: View the Avoiding Unfair Bias in Insurance Applications of AI Models - Chinese

Description: Every patient leaving the hospital has personal life-context circumstances that impact their chances of successful recovery and complications can be costly. Complex care management impacts over 50 percent of health care spend, with the admitted commercial population driving 37 percent of spend or $177M per 100,000 lives and the Medicare population driving 53 percent of spend or $600 million per 100,000 lives. The stakes are high, but existing care management solutions are mostly clinical and navigation call centers provide fragmented and outdated patient management systems, and do not address individual life-context. Discharged patients therefore remain underengaged and undercompliant, leading to costly complications. Innovative technology can help enable and integrate systems and existing infrastructure to leverage evidence-based contextualized care that will enhance care management, improve efficiency, and drive a better experience and outcomes for patients. This technology helps codifies life context, dynamically stratifying and engaging members digitally and telephonically helping virtual care teams to be more effective and efficient. In a recent randomized clinical trial, this technology and approach has shown a 50 percent reduction in the total cost of care, cutting the average and total cost of readmission, increasing virtual care team impact and member engagement and trust.

Description: This session introduces you to the next level of gen AI: AI Assistants (also known as Custom GPTs, Gems, or Bots). An AI Assistant is a specialized, tailored version of a Large Language Model (LLM) that allows the user to produce a persistent, pre-programmed tool designed to handle complex, repetitive professional tasks with guaranteed consistency, accuracy, and alignment. We will move beyond casual prompting to explore the mechanical architecture required to build reliable, high-stakes assistants for a variety of applications.