Combined Insurance, a Chubb company, is a leading provider of supplemental accident, health, disability, and life insurance products. Founded in 1922 and headquartered in Chicago, it operates at a significant scale (5,001-10,000 employees), serving individuals, families, and employers. The company's core function involves underwriting policies, managing a high volume of claims, and providing customer service for its benefits portfolio. As part of the global Chubb group, it benefits from extensive industry expertise while focusing on a specific niche within the broader insurance landscape.

Why AI matters at this scale

For a company of Combined's size and in the insurance sector, AI is not a futuristic concept but a pressing operational imperative. The business is fundamentally built on assessing risk, processing transactions, and managing relationships—all areas ripe for data-driven enhancement. At this employee band, manual processes for claims, underwriting, and customer inquiries create massive, repetitive workloads that are costly and prone to error. AI offers the leverage to automate these routine tasks, freeing thousands of employee hours for higher-value advisory and complex exception handling. Furthermore, in a competitive market for supplemental benefits, AI-driven personalization can be a key differentiator, helping customers choose the right coverage and improving retention. The scale of Combined's operations means that even marginal percentage improvements in efficiency or accuracy translate into millions of dollars in saved costs or increased revenue, providing a clear business case for strategic AI investment.

1. Automating High-Volume Claims Processing

The most immediate ROI lies in claims automation. Using natural language processing (NLP) and computer vision, AI can extract relevant data from submitted forms, medical documents, and even photos of incidents. This automates the initial triage, data entry, and validation steps, slashing processing time from days to hours or minutes. The impact is direct: reduced operational costs per claim, faster payment to customers (boosting satisfaction), and reallocation of human adjusters to complex, high-value cases that require empathy and nuanced judgment.

2. Enhancing Underwriting with Predictive Analytics

Underwriting supplemental policies involves evaluating applicant risk. Machine learning models can analyze a broader set of structured and unstructured data points—from application forms to external demographic and health trend data—to predict risk more accurately than traditional models. This allows for more precise pricing, reduces adverse selection, and can enable real-time, automated underwriting for standard cases. The financial return comes from improved loss ratios (more profitable books of business) and the ability to safely insure a broader population.

3. Deploying an AI-Powered Benefits Advisor

An intelligent chatbot or recommendation engine can guide employees through benefit selection. By analyzing an individual's role, family status, health history (with consent), and financial goals, the AI can proactively suggest the most relevant and valuable supplemental coverages. This improves the employee experience, increases uptake of appropriate products (driving revenue), and reduces the burden on HR and benefits administrators during enrollment periods.

Deployment risks specific to this size band

Implementing AI at a 5,000+ employee enterprise within a regulated industry like insurance carries distinct risks. The primary challenge is integration with legacy core systems (e.g., policy administration, claims platforms). These systems are often monolithic and not built for real-time AI model inference, making integration complex, expensive, and potentially disruptive to daily operations. Secondly, data quality and silos are a major hurdle. Relevant data may be scattered across departments, leading to incomplete training datasets for AI models. Third, regulatory compliance is paramount. Models used for underwriting or claims decisions must be explainable to meet state insurance regulations and avoid discriminatory biases, which can limit the use of some advanced "black box" AI techniques. Finally, change management is colossal. Successfully shifting the workflows of thousands of employees, from field agents to back-office staff, requires extensive training, clear communication, and demonstrating how AI augments rather than replaces their roles.