Zurich NA Head of Casualty on AI challenges & opportunities

David Morgan is the Head of Domestic Casualty for Zurich North America, where he is responsible for the strategy, business development and team focused on delivering tailored Casualty Insurance solutions for distributors and customers. As the insurance industry is undergoing a transformative journey as artificial intelligence (AI) technologies permeate various insurance lines, including casualty insurance, Morgan took time out recently to discuss the intricacies of AI's impact on casualty insurance.

Q: What are the biggest challenges facing customers seeking casualty insurance?

A: AI's ability to revolutionize the industry is undeniable, yet it's essential to address the primary hurdles. The foremost challenge is staffing. A shortage of skilled professionals, compounded by the need for rigorous training and retention strategies, looms large. Employee distractions also play a pivotal role. Losses are frequently attributed to distractions, staffing inadequacies, staff turnover, insufficient training, and the financial pressures burdening individuals. Recognizing and mitigating these challenges is vital for fostering sustainable growth.

Q: Can AI improve pricing or the availability of the most challenging casualty lines?

A: While AI’s prowess is impressive, it currently cannot replace the nuanced decision-making required for complex casualty lines. Instead, AI acts as a knowledge amplifier. With access to more information, AI enhances risk evaluation, facilitating better-informed pricing decisions. This may contribute to even more accurate pricing and greater certainty, albeit with the awareness that both upward and downward adjustments can occur based on refined insights.

Q: How can AI impact different types of casualty insurance, such as workers’ compensation?

A: AI's potential here is two-fold. It augments risk assessment, helping match employees to suitable roles based on behavior patterns and history. Additionally, AI's deepening understanding of individuals enables it to predict performance in various environments. This proactive approach not only could improve safety but also inform better risk assessment practices.

Q: What could be a potential pitfall of using AI for risk assessment?

A: Consider the concept of "black swan" events — rare and unexpected occurrences that can have catastrophic effects. They highlight that AI models, based on historical data, may not be able to predict such events accurately if they've never been encountered before. There are limitations of AI models in predicting the future. While AI can process historical data and trends, it might struggle to foresee events that deviate significantly from the past. There is the challenge of predicting events beyond a certain timeframe and the potential for the accumulation of concentration risks. There’s a need for insurers to take a balanced approach when considering the use of AI to help predict risks.

Q: Will insurance become obsolete due to AI's predictive capabilities?

A: No, AI will increase outcome predictability, which could help us get even better at matching pricing to the individual risk, but companies will still need protection for catastrophic events.

Q: How can Zurich responsibly use AI-driven information?

A: Balancing data privacy and compassionate assistance is key. Zurich can leverage AI for early issue detection, aiding in prevention. Ethical use of AI to improve lives while upholding insurance’s core values guides our approach.

Q: Can AI enable insurance companies to offer help to the most vulnerable individuals?

A: Absolutely. By identifying risks early and intervening, AI aligns with the compassionate values of insurance. It empowers companies to support vulnerable individuals, promoting healthier and safer lives.

Embracing AI's potential requires addressing challenges while navigating opportunities. As we venture into an era where AI transforms casualty insurance, it is paramount we strike a delicate balance between innovation and the ethical responsibility we have as human beings.