Insurance regulators on two continents are now explicitly asking financial institutions to maintain AI model inventories, a shift that moves AI governance from principle to operating practice. The National Association of Insurance Commissioners revised its draft AI Risk Evaluation Supplement this week to include direct inventory questions, while the Australian Prudential Regulation Authority published a letter telling banks, insurers, and superannuation trustees to document every AI tool, assign lifecycle ownership, map supplier dependencies, and prepare credible fallback plans for critical providers.
From principles to identifiable evidence
The NAIC supplement, revised after a 12-state pilot, now distinguishes between AI systems and AI models. It separates models with direct consumer impact from those with material financial impact and adds questions about explainability, materiality, and third-party model oversight. The document also introduces definitions for agentic AI, direct consumer impact, and material financial impact.
Regulators now explicitly ask for a model inventory. The supplement remains part of existing examination handbooks rather than a standalone regime, and examiners can limit their inquiries based on materiality and a company's initial responses. The direction is clear: a broad principle to govern AI responsibly is becoming a request for identifiable evidence about which models exist, what they affect, and who oversees them.
APRA reached a similar conclusion after reviewing large banks, insurers, and superannuation trustees. It reported that AI adoption was moving faster than governance, assurance, and operational-resilience practices. The regulator expects entities to maintain inventories of AI tools and use cases, preserve human involvement in high-risk decisions, map third- and fourth-party dependencies, and maintain credible substitution or exit plans for critical providers. APRA also found that point-in-time assurance was poorly suited to models that can change, drift, or degrade, and called for more continuous validation and monitoring.
For insurance compliance teams adapting to these expectations, structured training can help bridge the gap between policy and practice. AI Regulatory Compliance Courses offer frameworks for building the kind of documented governance programs these regulators now expect.
Finance leaders see returns, but process redesign lags
A survey of 102 CFOs and finance vice presidents at investor-backed companies, conducted by Consero Global, found that 97% reported using AI in the finance function and 76% said they had seen a return within 12 months. The sample is small and specialized, drawn from a firm that sells technology-enabled finance services, so the figures should not be read as a census of corporate finance or insurance.
The operational findings are more instructive. Many finance teams are adding AI to existing processes rather than redesigning the work. Data quality, accessibility, completeness, and siloed systems remain the leading obstacles. Consero also noted that measuring return through self-reported hours saved can be unreliable. Some teams are instead tracking concrete outputs such as a completed agent, automated workflow, or new dashboard.
An AI assistant that helps an employee complete the same process a little faster may produce value. A redesigned workflow that changes how evidence is collected, reviewed, approved, and recorded is a more substantial operating change and creates a better basis for measuring return. Adoption rates alone say little about operating maturity.
Who receives the economic gains
Anthropic released an interactive economic model built around three AI scenarios through 2030: modest, substantial, and extreme. These are scenarios, not forecasts. Depending on assumptions about AI capability, adoption, autonomy, productivity, and the time required for displaced workers to find new occupations, the model estimates that US GDP could be 1.6%, 8.3%, or 32.4% higher in 2030 than it would have been without AI.
The distribution of gains changes sharply with the scenario. In the substantial case, Anthropic estimates wages for knowledge workers would be essentially flat while demand and wages rise in less-exposed occupations. Labor's share of economic output falls from roughly 60% today to 56.1%. In the extreme case, knowledge-worker wages fall by more than 10%, unemployment rises sharply, and capital receives 54.8% of economic output.
Four years is a short period for workers to retrain and businesses to reorganize, and small changes in assumptions about adoption or job-transition time produce very different outcomes. The exercise shifts the question from "Will AI create growth?" to "How quickly can people, companies, and institutions adjust to the way that growth is created?"
Insuring the physical infrastructure
Marsh launched Stratus, a property insurance exchange for operational digital infrastructure, including data centers and related critical infrastructure. The facility provides access to as much as $10 billion of property insurance capacity in a single placement through approximately 30 capital providers.
That figure is placement capacity, not premium, and it does not mean every data-center risk can secure a $10 billion limit. Underwriting, engineering, geography, construction, power supply, cooling, equipment, and business-continuity dependencies still matter. The structure signals that AI companies and cloud providers are building facilities whose values and interdependencies can exceed the comfortable limits of conventional property programs. Marsh is organizing a broader pool of capital around operational risk rather than expecting a traditional placement to stretch indefinitely.
For CIOs and technology leaders managing the insurance implications of AI infrastructure, AI IT Strategy Training addresses the intersection of technology planning and operational risk that these new placement structures reflect.
Why this matters for insurance professionals
AI governance is acquiring an operating form. Regulators increasingly want evidence that an insurer knows what it has, what each system affects, who owns it, how it changes, and what happens if a critical provider fails. The inventory is becoming the control surface. At the same time, the economic scenarios from Anthropic suggest that insurance demand follows economic structure. A more capital-intensive economy could create large commercial opportunities around infrastructure, construction, energy, technology, and business interruption, even as payroll, employment patterns, and consumer demand change. The size of the economy is only part of the exposure. Its composition matters too. And as Marsh's $10 billion placement shows, the market is beginning to build new insurance architecture around the concentrated physical assets that AI depends on.
Your membership also unlocks:
