The National Insurance Crime Bureau has appointed Marta Magnuszewska (pictured) as senior vice president and chief information and technology officer, effective July 20.
In the role, she will oversee NICB's technology strategy and enterprise architecture, lead the information technology team across infrastructure, cybersecurity, application development and technical support, direct the data science team, and guide the organization's data governance, product development and AI direction.
Magnuszewska brings more than two decades of experience across the insurance industry, most recently serving as vice president of claims transformation, data analytics and AI at Markel Insurance Corporation, where she led enterprise-wide initiatives within the claims organization.
David J. Glawe, president and CEO of NICB, said Magnuszewska was a senior enterprise technology, data and analytics leader who had led modernization efforts spanning claims, underwriting and data ecosystems, including data pipelines, predictive models and analytics platforms that improved fraud detection, operational efficiency and claims outcomes.
"She has a proven track record of delivering measurable business impact, including scaling fraud analytics capabilities, modernizing claims operations through real-time data insights and integration, and leading global data and analytics strategies across multiple organizations," Glawe said.
The appointment comes as NICB and its member insurers face a persistent fraud problem alongside a newly active regulatory environment for the AI tools meant to fight it. NICB, citing Coalition Against Insurance Fraud data that isn't tied to a specific recent report, states that the United States loses an estimated $308.6 billion a year to insurance fraud, adding roughly $900 to the average policyholder's annual premium.
Supported by more than 1,200 property-casualty insurers, self-insureds, rental car companies, vehicle finance firms and auto auctions, NICB has flagged contractor fraud as a particular growth area, with reported cases rising 38% from 2023 to 2025 as severe weather events have grown more frequent and costly.
That scale of fraud is precisely why an AI-driven approach matters: manual investigation cannot keep pace with schemes that are increasingly organized and distributed across claims volumes this large. But fraud models carry a distinct risk alongside that promise. They often flag cases using proxies such as geography, claim history or contractor networks, and a poorly governed model that systematically over-flags certain populations becomes a discrimination and bad-faith claims-handling exposure for every insurer relying on it, not just a technical shortcoming.
Regulators are now checking for exactly that risk. Roughly half of US states, plus the District of Columbia, have adopted some version of the NAIC's Model Bulletin on the Use of Artificial Intelligence Systems by Insurers, though exact counts vary by tracker and reporting date. The bulletin requires carriers to maintain a documented AI governance program covering underwriting, rating, claims and fraud detection, along with audit rights over AI vendors and consumer disclosure obligations.
The NAIC has also begun piloting an AI Systems Evaluation Tool, a structured examiner questionnaire intended to assess insurer AI governance consistently, with wider rollout expected at the NAIC's 2026 Fall National Meeting.
That backdrop makes NICB's position unusually exposed compared with a single insurer building its own AI system. NICB functions as a shared industry utility, and its fraud flags, data products and analytics feed into decisions at hundreds of member insurers simultaneously.
If its AI infrastructure isn't defensible under examination, that risk does not stay contained to NICB; it propagates outward to every member insurer that acted on its output. A technology leader with direct experience building auditable, governed analytics systems, rather than standalone pilots, is therefore likely to matter to NICB's members as much as any efficiency gains the role delivers.
