AIG Leverages Generative AI to Transform Insurance Workflows
American International Group (AIG), a leading global insurance provider, has reported substantial advancements in operational efficiency and underwriting capacity thanks to its deployment of generative artificial intelligence (AI). During its recent Investor Day, the company disclosed measurable gains that surpass initial expectations, signaling a transformative shift in how AI is integrated into core insurance processes.
From Aspirational Goals to Concrete Results
Initially characterizing its AI-driven projections as “aspirational,” AIG’s CEO Peter Zaffino later acknowledged in a fourth quarter earnings call that the actual capabilities of generative AI exceeded these early benchmarks. “We’re seeing a massive change in our ability to process a submission flow way […] without additional human capital resources,” Zaffino commented, underscoring a notable increase in throughput without expanding staff.
Embedding AI in Underwriting and Claims
AIG has embedded its proprietary generative AI tool, AIG Assist, across most commercial lines of business, streamlining underwriting and claims procedures. The insurer’s excess and surplus unit, Lexington Insurance, aims to handle 500,000 submissions by 2030—already surpassing 370,000 in 2025 alone. This rapid scaling is enabled by generative models that extract and summarize incoming data efficiently.
Innovative Orchestration Layer Enhances Decision-Making
Central to AIG’s AI strategy is an orchestration layer designed to coordinate multiple AI agents. This system acts as a supervisor, driving better decision-making and reducing operational costs by integrating AI insights throughout the workflow. According to Zaffino, these AI agents function as “companions that operate with our teams,” providing real-time data, referencing historical cases, and challenging underwriting decisions to limit biases and accelerate processing times.
Streamlining the Front-to-Back Workflow
AIG emphasizes that the orchestration layer compresses the “front-to-back workflow,” tightening integration between submission intake, risk assessment, and claims handling. This coordination reduces repetitive tasks and shortens cycle times, enhancing overall efficiency.
Practical Applications in Portfolio Integration
The company has applied its AI stack to complex transactions, including the conversion of Everest’s retail commercial business, where account prioritization for renewal was achieved in a fraction of the usual time. By building an ontology combining Everest’s portfolio with its own, AIG improved portfolio blending—a technically demanding process that AI helped streamline.
Additionally, AIG partnered with Amwins, Blackstone, and Palantir to launch Lloyd’s Syndicate 2479, using large language models (LLMs) to evaluate portfolio alignment with defined risk appetites. Zaffino highlighted a “strong pipeline of SPV opportunities,” reflecting confidence in AI’s role in future strategic initiatives.
Implications for AI in Insurance and Beyond
AIG’s experience illustrates the tangible benefits of embedding generative AI within core insurance operations, showcasing how orchestration and workflow redesign can drive measurable economic impact. For AI decision-makers across industries, this case underscores the importance of integrating multiple AI agents and leveraging orchestration layers to boost capacity, reduce costs, and enhance decision quality.
Fonte: ver artigo original
Related Articles
Meta Commits to 100MW Solar Power for New AI Data Center in South Carolina
Amazon AWS Pledges Up to $50 Billion for AI and Supercomputing Expansion in U.S. Federal Sector
Samsung to Reveal Advanced Bespoke AI Kitchen Appliances at CES 2026
AI Awards 2026 Opens Submissions to Global Innovators in Artificial Intelligence
