AI Insights Feb 27 2026

Feb 27, 2026

insuranceindustry.ai article hero image

AI Insights: February 27, 2026

Your weekly analysis of AI developments in insurance.

Anthropic’s CEO Says the Tsunami Is Already Visible

Dario Amodei, the CEO of Anthropic, one of the leading AI development companies, appeared on a podcast this week with Zerodha co-founder Nikhil Kamath and said something that insurance executives should take seriously: society is not prepared for what is coming, and the window to prepare is shorter than most people think.

Amodei described the current moment as watching a tsunami approach from the shore. “It’s so close, we can see it on the horizon,” he said, “and yet people are coming up with explanations like, ‘Oh, it’s not actually a tsunami, it’s just a trick of the light.'” He argued that AI is reaching human-level capabilities faster than most people realize, and that the economic and geopolitical disruption it will cause is not a distant possibility but an imminent one.

He also said AI’s impact will vary by industry depending on the nature of the tasks involved, noting that while AI models are already handling a significant and growing share of coding work, broader responsibilities will take more time to automate. He is optimistic about healthcare outcomes, suggesting AI could accelerate the cure of diseases that have resisted human medicine for generations.

Why This Matters:

Amodei is not a commentator. He runs the company that built Claude, one of the most widely deployed enterprise AI systems in the world, and that has partnership agreements with major carriers including Travelers, Allianz, and others. When he says the wave is visible and most people are explaining it away, he is describing a pattern he watches from a position very few people occupy.

The tsunami metaphor is particularly worth sitting with. A tsunami doesn’t announce itself with noise as it builds. It draws the water back first. What the insurance industry has been watching for several years — the incremental pilot programs, the proof-of-concept deployments, the cautious board discussions — may be the waterline receding. What comes next is not incremental.

The insurance industry has already seen what Amodei is describing at the margins. Travelers built a complete agentic voice claims system in a matter of months. ERGO announced a thousand job reductions tied to AI automation. The Gallagher Re report this month documented $5 billion in AI-focused insurtech investment last year alone. These are not individual data points. They are a pattern.

Strategic Implications:

Amodei’s comments are most useful not as a warning to panic but as a calibration tool. If the person leading one of the most advanced AI development organizations in the world believes the transition is happening faster than most institutions recognize, that belief should inform how much urgency your organization brings to AI readiness.

The carriers that will be best positioned when the full force arrives are those that have treated AI preparation not as a future-year budget item but as a current operating priority. That means building internal expertise now, not hiring it later. It means moving from pilot programs to production deployments on a compressed timeline. And it means taking seriously the governance infrastructure that will determine whether your AI implementation creates competitive advantage or regulatory liability.

The tsunami is visible. The question is what you are doing about it while you can still see it coming.

A ChatGPT App Crashed Broker Stocks. Now What?

When Insurify launched an insurance shopping app in OpenAI’s ChatGPT directory on February 3, the market reaction was immediate and severe. The S&P 500 Insurance index dropped 3.9% on February 9, its largest single-session decline since October. Willis Towers Watson closed 12% lower — its worst trading session since November 2008. Arthur J. Gallagher fell 9.9%. Aon fell 9.3%.

That drop triggered a great deal of breathless coverage about AI disrupting the insurance brokerage industry. Insurance Journal devoted its February 23 cover story to examining the disruption question seriously. The conclusion is more nuanced than the market’s initial reaction suggests, but it is not reassuring.

The Insurify app, the first insurance app in OpenAI’s directory, lets consumers compare auto insurance rates within a ChatGPT conversation. Users describe their situation in plain language, receive tailored estimates from carriers in their area, compare coverage options and customer service ratings side by side, and transition to Insurify’s licensed platform to bind the policy. Insurify CEO Snejina Zacharia framed it as making insurance shopping feel like having a conversation rather than filling out forms.

Why This Matters:

The market’s reaction was emotional and probably overdone in the short term. A ChatGPT plugin for auto insurance comparison is not going to displace major commercial brokers immediately. But the fear driving that 9.9% drop at Gallagher is not irrational. It is a recognition that the distribution model for personal lines, which depends on friction and consumer confusion to create value, is exactly the kind of model that AI-native interfaces erode.

Insurify is a licensed agent in all 50 states and Washington, D.C. It is not a technology company pretending to sell insurance. It is a digital agent that has now placed itself inside the most-used AI consumer interface in the world. Experian followed suit within weeks, launching its own insurance marketplace app inside ChatGPT. The entry point for insurance distribution is shifting.

The commercial and wholesale broker market operates on relationships, expertise, and complexity that does not reduce to a consumer chat interface. That segment faces a different kind of AI disruption — the operational and underwriting AI that this newsletter has covered for months, not disintermediation through a shopping plugin. But personal lines distribution is a different story, and the market recognized that even if it overshot in the immediate reaction.

Strategic Implications:

For independent agents operating in personal lines, the Insurify development is not theoretical. A consumer who can get accurate rate comparisons across carriers inside a ChatGPT conversation, without calling an agent or filling out a form, has less reason to initiate that call. The question is not whether this creates pressure on traditional personal lines distribution. It does. The question is how agents reposition the value they provide.

The answer is not competing on quote speed or comparative shopping, where AI will win. It is competing on advice, relationship, and complexity — the coverage gaps a consumer comparison tool does not catch, the claims advocacy that matters when something goes wrong, and the relationship with a licensed professional who knows the client’s situation in ways a chat session does not. That is a genuine value proposition. But it requires agents to articulate it, invest in it, and lead with it rather than treating the comparison function as their core differentiator.

Harper Raises $47 Million to Build an AI-Native Brokerage

An AI-native commercial insurance brokerage called Harper announced a $46.8 million combined seed and Series A round on February 25, led by Emergence Capital and backed by Y Combinator and Peak XV Partners. The company has raised $54 million in total since launching in 2024, doing so in under two months.

Harper’s founder, Dakotah Rice, comes from a family that owned an insurance brokerage. He built the company specifically because he remembered how painful and slow the commercial insurance process was for the small and mid-sized businesses that walked through his family’s doors. His pitch is straightforward: what takes a traditional broker five to seven days, Harper does in one to two. A typical human-led brokerage sales team handles 20 to 30 deals per month. Harper says it handles more than 1,000 customers per month.

The model is built around AI handling submission routing, underwriter follow-ups, document collection, and pipeline management, with licensed human oversight at the points where it matters. Harper matches small and mid-sized businesses with more than 160 carriers across workers’ compensation, general liability, and professional liability. The company has over 5,000 customers. The new funding will go toward expanding the engineering team and building the brand.

Rice’s vision extends beyond insurance: he wants Harper to become the operational backbone for entrepreneurs, handling risk, compliance, and back-office functions over time. The insurance relationship is the entry point, not the ceiling.

Why This Matters:

Harper is not the only AI-native brokerage in this space, but the amount it raised and the investors behind it are signals worth paying attention to. Emergence Capital has a track record in enterprise software. YC has been publishing thinking about agency distribution that aligns with Rice’s model. The thesis — that insurance agencies will increasingly look like software companies, with software margins — is gaining institutional backing.

The traditional commercial brokerage serving small and mid-market accounts is particularly exposed to this model. The complexity that protects large-account brokers from AI disintermediation is not always present in the small commercial segment. Routine workers’ comp, GL, and professional liability placements involve significant process work that AI handles efficiently. If Harper and companies like it can automate that process while maintaining placement quality and carrier relationships, the economics of the traditional small commercial brokerage become difficult to defend.

Strategic Implications:

For established agencies and brokers in the small commercial market, Harper’s raise is a competitive signal. The question is not whether AI-native brokerages will take market share. They will. The question is what your agency is doing to compete on the dimensions where a software company cannot match you.

For wholesalers and MGAs, the Harper model is also worth watching from a distribution perspective. A brokerage that handles over 1,000 customer accounts per month and routes submissions through AI is a different kind of distribution partner than a traditional retail agent. The submission quality, documentation, and data that come through an AI-native channel may actually be cleaner and more consistent than what arrives from a manual process. That creates both opportunity and a pricing challenge.

The NAIC Puts AI Governance at the Top of Its 2026 Agenda

The National Association of Insurance Commissioners released its strategic priorities for 2026 this week, and for the first time, AI governance sits in a named category alongside disaster preparedness and capital frameworks. The NAIC called it “Leading on AI Model Governance, Innovation Oversight, and Cyber Threats,” and the language it used is worth reading carefully.

The NAIC’s position is that the state-based regulatory system is the appropriate framework for AI oversight in insurance, and it intends to defend that position actively. That matters because it is operating in direct tension with a federal AI executive order signed in December 2025 that the NAIC has publicly opposed, saying it creates significant unintended consequences and legal uncertainty for insurance markets.

The practical regulatory activity moving forward in 2026 is substantial. Several states, including Florida, Connecticut, Pennsylvania, and Iowa, are preparing to pilot the NAIC’s AI Systems Evaluation Tool during insurance examinations this year. The tool is designed to standardize how regulators assess insurers’ AI governance, and its use in actual examinations marks the transition from guidance to enforcement. The NAIC’s Big Data and AI Working Group held a public meeting on February 17 to continue developing the tool, and its next in-person session is scheduled for March.

Twenty-four states have adopted the NAIC’s AI Model Bulletin as of early 2026, requiring insurers to maintain documented AI governance programs covering transparency, bias testing, fairness, and vendor oversight. Colorado’s broader AI Act took effect February 1. The regulatory framework is no longer theoretical.

Why This Matters:

The NAIC’s declaration of AI governance as a top 2026 priority, combined with the active piloting of an examination tool, means that AI governance is moving from a voluntary best-practice discussion to a mandatory compliance function. Carriers that have treated their AI programs as informal arrangements — no documented governance, no vendor oversight structure, no bias testing protocols — are now operating with exposure.

The federal-state tension over AI regulation creates an additional layer of uncertainty that carriers should not mistake for a slowdown. The NAIC has been explicit that it will not defer to federal preemption efforts on insurance AI regulation, and it has the institutional credibility and state-level authority to hold that position. The patchwork of state requirements that results from the NAIC’s decentralized model is genuinely complex for multi-state carriers, but it is not going away.

The AI Systems Evaluation Tool is a particularly important development to watch. When regulators pilot a new examination tool, they learn what they are looking for. The carriers that volunteer for or participate early in that process will have a significant advantage in understanding what documentation and governance structures satisfy regulatory review. Those that wait will find out during their own exam.

Strategic Implications:

For chief compliance officers and risk management teams, the 2026 NAIC agenda is a planning document. The AI governance requirements that are now expected include a documented AI program covering all AI systems in use, ongoing bias testing for underwriting and claims models, a third-party vendor management framework that includes oversight rights and cooperation requirements, and audit-ready documentation of governance decisions.

For carriers that have already built these capabilities, the regulatory environment provides a competitive signal. An insurer with demonstrable AI governance infrastructure can move faster on new AI deployments because the governance scaffolding is already in place. Carriers that have not built this infrastructure will find themselves building it reactively, under regulatory scrutiny, while trying to keep up with competitors who are moving forward.

The bottom line is this: the NAIC’s 2026 priorities make clear that AI governance is now a compliance function, not a best-practices aspiration. The examination pilots beginning this year will clarify exactly what that means in practice.

The Bottom Line

This week’s stories share a common thread: the distance between awareness and action.

Dario Amodei can see the tsunami. He leads one of the companies building it, and he is describing an institution-level failure of preparation. The Insurify story shows what early disruption looks like at the distribution level — a visible demonstration that AI is restructuring the consumer’s path to insurance purchase, right now, not eventually. Harper’s raise shows that well-capitalized, AI-native competitors are building the infrastructure to take market share in commercial lines while established players are still debating strategy. And the NAIC is now saying, in plain language, that AI governance is a 2026 compliance priority — meaning the cost of inaction is no longer just competitive, it is regulatory.

The insurance industry has advantages in this transition that most industries do not have. It has deep proprietary data. It has regulatory relationships that create barriers to entry. It has customer relationships built over decades. And it has a product that consumers genuinely need, regardless of how the distribution and processing of that product evolves.

Those advantages are real. But they are not infinite. They erode when AI-native competitors build equivalent data relationships, when regulators develop frameworks that apply equally to incumbents and new entrants, and when customers discover that the friction they accepted was not the same as the value they needed.

The carriers and agencies that will lead through this transition are the ones treating urgency as a competitive asset, not a source of anxiety. Act now, while the wave is still visible.

AI Insights appears every Friday, analyzing AI developments through an insurance lens. For deeper analysis of strategic implications, visit InsuranceIndustry.ai.

By James W. Moore

Sources:

Subscribe to AI Insights Newsletter

AI Disclaimer: This blog post was created with assistance from artificial intelligence technology. While the content is based on factual information from the source material, readers should verify all details, pricing, and features directly with the respective AI tool providers before making business decisions. AI-generated content may not reflect the most current information, and individual results may vary. Always conduct your own research and due diligence before relying on information contained on this site. 

Related Posts

468