The signals this month are pointing in the same direction. AI is separating the industry into two groups — and the line isn’t between early adopters and laggards. It’s between organizations that are building AI into their operating layer with real governance and those that are bolting it on and hoping for the best. The data on who’s winning is now unambiguous. The data on who’s failing — and why — is equally clear. Meanwhile, deepfake fraud is changing the claims conversation in ways most agencies aren’t prepared for, and the M&A market is quietly raising its standards for what a sellable agency actually looks like.
1. AI governance gap is killing insurance AI projects
Grant Thornton’s 2026 AI Impact Survey Report surveyed 100 insurance executives on how AI is performing inside their organizations. The results show a split that should concern anyone building an AI strategy right now. On the positive side, 52% of respondents credit AI with revenue growth, 62% report better decision-making, and half say it has already cut costs. On the negative side, 44% say governance or compliance challenges have contributed to AI projects failing or underperforming — and only 24% say they are very confident they could pass an independent AI governance review in 90 days.
The most common failure mode is fragmentation: 61% of executives say their boards have established AI governance policies, but the evidence and controls to back those policies are scattered across teams and tools rather than centralized and auditable. Grant Thornton’s framing is direct: without tested governance, insurers are leaving themselves open to regulatory risk, customer risk, and financial pressure that erodes the very ROI they were chasing. The firms that have invested in governance infrastructure are now deploying AI into higher-value workflows — underwriting, claims, pricing — and compounding returns. The ones that skipped governance are stuck managing AI failures in the back office.
Key takeaways:
- The AI governance gap isn’t a future problem — it’s the reason nearly half of insurance AI projects are already underperforming. The technology is available. The operational discipline to deploy it safely is not.
- Only 24% of insurance executives could confidently pass an independent AI governance review in 90 days. For regulators who are increasingly scrutinizing automated decision-making in underwriting and claims, that number represents significant exposure.
- The pattern among high-performing firms is consistent: governance first, then deployment into higher-value workflows. Agencies and carriers that treat governance as a checkbox will keep getting the same results they’re getting now.
Source: Claims Journal / Grant Thornton 2026 AI Impact Survey Report
2. Deepfake fraud is now hitting every line of business
IA Magazine’s May report documents something carriers have been tracking quietly for months: AI-generated and manipulated content is now showing up across personal lines, commercial lines, and specialty coverages in insurance claims. The specific examples are instructive — inflated home damage photos, staged car accident documentation, synthetic medical scans, and AI-manipulated images of luxury items used to inflate personal property claims. Some of these are not sophisticated fraud rings. They are policyholders filing legitimate claims and using widely available AI image tools to exaggerate damages. Verisk has separately flagged that traditional review methods were not designed to catch synthetic media, and the gap between what fraudsters can generate and what adjusters are trained to detect is widening.
The part most relevant to independent agencies: IA Magazine and Verisk both say independent agents now have an active and growing role in fraud detection — particularly for larger commercial and personal property losses — by directly verifying damage, contractor estimates, and repair documentation with insureds before those claims move through the system. The agents whose relationship with the insured runs deeper than a policy renewal date are the ones who can actually perform that verification function credibly.
Key takeaways:
- Deepfake fraud is no longer a carrier problem that stays in the claims department. It’s moving into the agency relationship — and the agencies that are close to their insureds have a real role to play in catching it before it costs everyone.
- The availability of AI image tools means the volume of this type of fraud will keep rising regardless of carrier detection investment. Agents who build verification into their post-loss workflow are adding measurable value to their carrier relationships.
- Traditional review processes weren’t designed for synthetic media. Agencies and carriers that update their workflows and training now will be in a better position when this becomes a standard underwriting and claims concern — which it will.
Source: IA Magazine — Deep Fakes: The New Front Line in Insurance Fraud
3. Insurance M&A in 2026: a selective market doesn’t mean a closed one
Brown & Brown’s acquisition team published one of the more candid assessments of where the agency M&A market actually stands heading into mid-2026. The core argument: volumes have slowed and valuations have likely hit a high-water mark, but buyers are still active. What has changed is what they’re looking at. Several of the highest-volume acquirers from the roll-up era have themselves been acquired, and the newer buyers entering the market don’t carry the same institutional depth or long-term mindset. For sellers who care about what happens to their team and their clients after the deal, the pool of buyers who share that mindset has genuinely narrowed.
The Brown & Brown piece is direct on what stands out in the current environment: strong leadership, a sales team that isn’t dependent on the owner, differentiated capabilities or geography, and — critically — operational quality that is demonstrable and repeatable under scrutiny. Buyers are spending more time observing how decisions get made day-to-day and less time chasing momentum on top-line premium growth. Surface-level numbers matter less than durability. The advice for sellers thinking about the next two to three years is to treat preparation for a sale with the same discipline as growing the business: build something resilient and intentional, not something that looks good for a few months before a meeting.
Key takeaways:
- The window hasn’t closed, but the criteria have shifted. Premium volume alone isn’t what gets the best multiples anymore. Operational quality, leadership depth, and book durability are the variables buyers are actually underwriting.
- The pool of buyers who genuinely care about staff continuity, client continuity, and cultural fit has narrowed as institutional roll-up buyers have been absorbed by larger platforms. Sellers who care about those things need to be more selective about who they talk to — not less.
- The agencies that will command the best outcomes over the next two to three years are the ones investing now in operational discipline, clean books, and reduced owner dependency. That work takes time — and it’s the same work that makes the business worth owning whether or not a sale ever happens.
Source: Brown & Brown / Arrowhead Programs — Insurance M&A in 2026
4. Capgemini: AI “intelligence trailblazers” are posting 21% higher revenue growth and 51% greater share price gains
Capgemini’s 19th annual World P&C Insurance Report is one of the more rigorous annual benchmarks in the industry — drawing on 344 senior executives, 809 employees, and 1,113 policyholders across the Americas, Europe, and Asia-Pacific. The standout finding this year is the performance gap between what Capgemini calls “intelligence trailblazers” — carriers treating AI as a core operating capability rather than a technology initiative — and the rest of the market. The gap is not marginal. Trailblazers are achieving up to 21% higher revenue growth and 51% greater share price gains over three years compared to their peers.
The report defines the trailblazer profile not by which AI tools a carrier has deployed but by how deeply AI is integrated into operating decisions, risk workflows, and customer experience. It is an architecture story as much as a technology story. The report also flagged a pattern consistent with Grant Thornton’s governance findings: the carriers seeing the strongest returns are the ones that built operational discipline and governance infrastructure before scaling AI across higher-value functions. The ones chasing AI adoption without that foundation are generating activity, not returns.
Key takeaways:
- A 21% revenue growth advantage and 51% share price premium over three years is not a pilot program result. It is the compounding effect of treating AI as an operating layer rather than a set of tools. That gap will widen as trailblazers keep deploying into higher-value workflows.
- The trailblazer profile isn’t defined by the tools deployed — it’s defined by how deeply AI is integrated into operating decisions. Two carriers can have identical AI vendors and produce completely different outcomes depending on whether the operating model was redesigned around the capability or just layered on top of the old one.
- For agencies evaluating carrier relationships and technology partners, this is the benchmark to watch. The carriers investing in AI as an operating capability now are the ones that will have the most competitive capacity, the most disciplined underwriting, and the most stable appetite as the market continues to evolve.
Source: Roots.ai / Capgemini 19th Annual World P&C Insurance Report