AIG Harnesses Generative AI to Transform Insurance Processes
American International Group (AIG), a leading global insurance company, has reported faster than anticipated progress in integrating generative artificial intelligence into its core business operations. This development signals notable improvements in underwriting capacity, operational efficiency, and portfolio management.
Measurable Impact on Throughput and Workflow Redesign
During a recent Investor Day presentation, AIG shared insights into the measurable gains achieved through its AI initiatives. While CEO Peter Zaffino initially described early expectations as aspirational, he later acknowledged in a quarterly earnings call that the capabilities of AI exceeded those projections. He emphasized a “massive change” in processing submission flows without increasing human capital, highlighting AI’s role in driving operational scalability.
Embedding AI into Underwriting and Claims
AIG has embedded generative AI extensively within its underwriting and claims workflows through its proprietary tool, AIG Assist, which is now implemented across most commercial lines. The insurance giant’s excess and surplus unit, Lexington Insurance, aims to process 500,000 submissions by 2030 and has already surpassed 370,000 in 2025, reflecting accelerated processing capacity enabled by AI.
Introducing an Orchestration Layer for AI Agents
One of the most significant innovations is AIG’s development of an orchestration layer within its AI technology stack. This layer coordinates multiple AI agents, enabling them to collaborate effectively to enhance decision-making and reduce organizational costs. Unlike prior periods where orchestration was less emphasized, this approach tightly integrates AI capabilities across the entire underwriting workflow.
CEO Zaffino describes these AI agents as “companions” working alongside human teams, providing real-time data, referencing historical cases, and challenging underwriting decisions to improve accuracy and efficiency. This orchestration reduces processing times dramatically and ensures unbiased data analysis throughout the workflow.
Streamlining Front-to-Back Workflows
AIG’s orchestration strategy compresses the so-called “front-to-back workflow,” integrating key stages such as data intake, risk assessment, and claims handling. Coordinated AI agents automate repetitive and time-consuming tasks, enabling faster and more consistent outcomes.
AI in Portfolio Integration and Special Purpose Vehicles
AI’s impact extends to strategic transactions. During the acquisition of Everest’s retail commercial business, AIG employed AI to prioritize account renewals rapidly by building an ontology that aligned Everest’s portfolio with AIG’s own. Ontological alignment, a complex technical process, was successfully leveraged to blend portfolios efficiently.
In partnership with Amwins and Blackstone, AIG launched Lloyd’s Syndicate 2479, applying a similar ontological and AI-driven approach to assess risk alignment using large language models (LLMs). Collaborating with Palantir, this initiative reflects AIG’s strong pipeline of special purpose vehicle (SPV) opportunities enhanced by AI.
Implications for AI Decision-Makers
AIG’s experience highlights how embedding generative AI within core insurance processes, combined with an orchestration layer for agent collaboration, can deliver measurable economic benefits. The company demonstrates that increased capacity and reduced cycle times are tangible outcomes, providing a compelling case study for AI adoption in complex, data-intensive industries.
(Image source: “Nagasaki, AIG (Insurance company) building” by Admanchester, licensed under CC BY-NC-ND 2.0.)
Related Articles
Telehealth Startup Medvi’s $1.8 Billion Revenue Fueled by AI-Driven Fake Advertising
How Founders Can Differentiate Their AI Startups in a Saturated Market, Says VC Jennifer Neundorfer
Salesforce Unveils Advanced Slackbot AI to Rival Microsoft and Google in Enterprise AI
YouTube Introduces AI-Powered Guided Search for Premium Users in the U.S.
