AI Used to Automate Insurance Underwriting
Actuarial and underwriting teams are increasingly adopting AI to automate data ingestion and streamline risk assessment. Intact Insurance's change director described using AI-powered data extraction to increase efficiency and speed for brokers. The focus remains on integrating AI with human expertise to augment decision-making rather than fully replacing underwriters.
- Insurers are leveraging MLOps platforms to overcome the challenges of scaling machine learning models from experimentation to production; a key focus is on continuous monitoring to prevent issues like "model drift" and ensure models perform as expected. The modern data stack for these initiatives often includes tools like Snowflake for cloud data warehousing, Airflow for workflow orchestration, and dbt for data transformation. - Actuarial bodies are establishing governance frameworks for AI to address risks like algorithmic bias and model explainability. The International Actuarial Association has released guidance on the testing, documentation, and governance of AI models to ensure transparency and accountability in risk assessment. - Building effective AI engineering teams involves creating specialized roles beyond data scientists, including ML Platform Engineers who build internal tools and AI Implementation Engineers who focus on deploying existing models into production systems. Effective AI teams often start small (2-3 engineers) and establish clear protocols for technical design reviews and documentation to prevent knowledge silos. - Liberty Mutual reduced its commercial insurance underwriting time by 40% using AI to assess risk factors and suggest policy terms from real-time data. Similarly, Zurich Insurance implemented an AI underwriting engine that cut its policy processing time by 50%. - In consumer tech, AI is being used for hyper-personalization in fashion retail. Companies like Stitch Fix use a combination of generative AI and human stylists to provide tailored clothing recommendations, while Zara applies machine learning to optimize inventory levels based on demand forecasts, reducing overproduction. - Big tech companies are pushing AI into new hardware. Apple is developing several AI-powered wearable devices, including smart glasses and AirPods with expanded AI capabilities built around a new version of Siri powered by Google's Gemini models. Meanwhile, OpenAI has hired a large team to develop its own consumer hardware, starting with a smart speaker. - New York City's AI startup scene is expanding, with a focus on enterprise, fintech, and health tech AI. Companies like Cyera (AI-powered data security) and Tennr (AI automation for medical documents) have recently raised significant funding rounds and are hiring for roles including Machine Learning Engineer and AI Product Manager.
