Independent Analysis of AI in Insurance

A curated directory of the regulators, public filings, research institutions, actuarial organizations, ratings agencies, standards bodies, and industry publications behind serious insurance research.

AI is becoming institutional infrastructure. This week: Peter Zaffino joins Palantir, the UAE defines authority for agentic AI, researchers find hidden heart disease in routine ECGs, and Anthropic’s reported revenue run rate reaches $65 billion. AI Insights Sep 4 2026

An independent directory of AI integration companies serving insurance, including implementation firms, data integration providers and AI orchestration platforms. Explore documented capabilities for connecting artificial intelligence to policy administration, claims systems, underwriting workflows and enterprise data. Listings are organized alphabetically by category, with source links and insurance-specific context.

Insurance has spent decades automating underwriting by defining where machine authority ends. Generative AI changes what machines can do, but it does not change who owns the decision architecture surrounding evidence, rules, authority, judgment, and the record.

This week’s AI developments point to a broader shift for insurance: the model itself may be becoming less important than the system built around it. NVIDIA’s agent architecture, Meta’s workforce experiment, MIT’s extreme-event research, and thousands of filed AI exclusions all show what changes when artificial intelligence moves from model capability into real operating environments.

Large language models are often criticized for being non-deterministic, but human insurance decisions have never been deterministic either. Research found a 55% median difference between underwriters pricing the same cases. AI changes the problem by making variance more measurable and consistent, but shared models can also turn independent errors into correlated ones. For insurers, the real question is not whether AI is deterministic, but what shape of error they want on the book.

AI is moving closer to the messy reality of insurance operations. This week: agentic underwriting workflows, OpenAI slowing model development over cyber concerns, Munich Re’s At-Bay acquisition, and the growing push to define AI-related coverage.

If AI makes carriers better at selecting and pricing risk, should they need less reinsurance?

Probably not. Reinsurance demand is shaped by more than underwriting confidence, including capital, liquidity, volatility, concentration, and price. And if carriers increasingly rely on similar models and decision systems, better individual risk selection could actually make diversification harder.

Part 3 of The Economics of AI in Insurance looks at what AI may change for reinsurance, and what it probably won’t.

AI is moving from promise to operating reality. This week: regulators sharpen examination tools, Allianz prices transformation, data centers strain insurance capacity, and reinsurers confront the data problem.

Everyone expects AI to cut underwriting costs. Almost no one is asking where the resulting capacity goes. This thought experiment follows it, and finds the more important question hiding underneath: not what AI does to expenses, but what it does to uncertainty.