How AIG Is Reinventing Insurance With AI | Peter Zaffino

Peter Zaffino arrived at AIG when its core property‑and‑casualty unit had amassed about $33 billion of underwriting losses from 2008‑2018. Rising from COO to president, CEO, and now executive chairman, he was tasked with stopping the cash bleed and modernizing operations. A cornerstone of his strategy is a bold AI initiative that embeds machine‑learning models into underwriting and claims, delivering faster risk assessment, lower manual effort, and a clear competitive advantage.

AIG’s AI engine combines deterministic, probabilistic and stochastic techniques to gauge catastrophe exposure worldwide. Earthquake insurance in California, for instance, is priced by aggregating exposure, applying probability curves, and capping limits to safeguard the balance sheet. Similar analytics steer hurricane, wildfire and tsunami underwriting, giving underwriters real‑time risk metrics. Automated data feeds from satellite imagery, sensor networks and historic loss records accelerate scenario testing, helping actuaries refine pricing while staying compliant.

Reinsurance acts as AIG’s safety net, with billions of dollars of treaty and facultative cover from a select group of global partners who understand the firm’s risk appetite. AI‑driven insights let reinsurers price capacity more precisely, lowering protection costs and freeing capital for growth. Lessons from 9/11 and the 2008 crisis have hardened AIG’s risk culture, emphasizing diversified work sites, robust continuity plans, and scenario‑based stress testing. Zaffino sees an insurance future where predictive analytics, automated claims and agile governance drive sustainable profitability.