Where Commercial Insurance AI Capital Landed in 2026
Q1 2026 insurtech funding hit $943M with 75% flowing to AI-native platforms. Here is where the money went and what a $50M broker should do about it.
TL;DR. Global insurtech funding hit $943 million in Q1 2026, up 27% year over year, with roughly 75% flowing to AI-native platforms. Federato closed $100M, Akur8 has now raised $180M cumulative, Honeycomb pulled $40M, Novella raised $21M, Comeryx launched with $7.5M. The investment thesis is clear: AI is rebuilding the commercial insurance stack. The deployment gap is the story underneath. 61.3% of MGAs report using AI today. Only 35.5% have a formal AI budget. If you run a $40M to $80M brokerage, MGA, or program administrator, the question in 2026 is not which vendor to evaluate. The question is whether you have a line-item budget at all.
If you run a $50M wholesale brokerage or program administrator, you have seen the inbox traffic. Federato. Akur8. Cytora. Novella. Honeycomb. Comeryx. Six AI-native commercial insurance platforms have closed funding rounds in the last twelve months, four of them in the last six. Total capital deployed on the visible rounds: north of $300 million. The conversations are louder, the demos are sharper, and the carriers you place with are quietly rolling out AI underwriting tools your retail agents are starting to ask about.
You have probably trialed one of them. You have probably told the board "we are looking at AI" in your last quarterly review. And you almost certainly do not have a formal AI budget that made it through the last planning cycle.
That gap, between using AI and funding it, is the real story of where commercial insurance capital landed in 2026.
The capital flow, by the numbers
Global insurtech investment reached $943.4 million across 42 deals in Q1 2026, a 27% year-over-year increase, according to Gallagher Re's quarterly InsurTech intelligence report. Approximately 75% of that capital went to companies with AI at the core of the product, a share that has roughly doubled since 2023. The shift is structural. In a softening commercial lines rate environment, carriers and intermediaries are turning to AI not for top-line growth but for cost-side leverage, and capital is following.
The single largest commercial-lines AI round of the last twelve months was Federato's $100 million Series D in November 2025, led by Growth Equity at Goldman Sachs Alternatives. The round took Federato's total funding past $180 million. The company's pitch is direct: agentic AI that handles submission triage, appetite matching, and quote generation in seconds, freeing underwriters to focus on judgment and relationships. Federato more than tripled revenue in the twelve months before the round. Its customers span carriers and high-growth MGAs.
Akur8, the AI-native non-life pricing and reserving platform, has reached the same $180 million cumulative funding mark on the back of its $120 million Series C and the recent acquisition of the Arius reserving platform. Akur8 now serves more than 250 customers across 40 countries. Its claim: actuaries cut pricing model build time from weeks to hours, and the reserving integration closes a loop that legacy systems leave open.
Honeycomb Insurance raised $40 million in June 2026 to expand its AI underwriting platform for apartment buildings and condo associations. Honeycomb exited 2025 with $275 million in gross written premium. Its model is no-inspection underwriting on properties evaluated through high-resolution imagery, geospatial datasets, and proprietary risk models.
On the brokerage side, Novella raised $21 million in May 2026 to expand its AI-powered wholesale operation. Founded in 2024, Novella now holds appointments with close to 100 E&S carriers, runs business through 3,500 retail agencies, and targets $10 billion in premium volume within a decade. Their stated goal is to become one of the top five wholesale brokers in the United States.
And Comeryx launched in February 2026 with a $7.5 million seed led by Altai Ventures, with Arch Capital, American Family Ventures, and Intact Ventures participating. Comeryx is building an AI-native MGA for the small-business artisan contractor market, the sub-$10,000 premium accounts that wholesale brokers usually walk past because manual underwriting math does not work.
That is roughly $288.5 million of disclosed AI-native commercial insurance funding in twelve months, concentrated in three workflow categories.
Where the money is concentrating
The capital is not spread evenly across "insurance AI." It is concentrating in three narrow lanes where the unit economics already work.
Submission intake and risk triage. Companies like Cytora, Novella, Convr, and Send are rebuilding the bottom of the funnel. Submissions arrive as PDFs, emails, and ACORD forms. AI ingests them, extracts the structured data, runs appetite and pricing rules, and surfaces only the in-appetite, complete submissions to a human underwriter. McKinsey reports that algorithmic triage is boosting underwriting capacity by roughly 50% and processing submissions 5x faster. Cytora's client Zurich rolled out AI risk digitization across five countries in 90 days, reducing manual processing time by more than 80%.
Underwriting decision support and pricing. Federato and Akur8 own this lane. The pitch is not automation. It is leverage: the underwriter still owns the decision, but the AI delivers a complete, explained, on-strategy quote in minutes instead of days. The same underwriter team can write twice the business at a tighter combined ratio because the analysis stops being the rate-limiter.
AI-native MGAs. This is the most aggressive bet. Instead of selling tools to existing MGAs, Comeryx and Honeycomb are building entire MGAs around AI underwriting from the founding stack up. The wager: in segments where manual underwriting economics fail (sub-$10K artisan contractor premiums, low-margin landlord property), an AI-native MGA can profitably write business that incumbents cannot.
Notably absent from the top of the league table: AI for personal lines, AI for health and life, and broker-facing chatbots. McKinsey and the Q1 2026 funding data both show those categories trailing. Capital is voting for commercial P&C workflows where dollars per submission justify the integration cost.
The path you are probably on
If you run a $50M brokerage or MGA, your AI journey almost certainly looks like this. A producer asked about ChatGPT in early 2024. Someone on the team signed up for a Cytora demo. You ran a 90-day pilot on submission triage in 2025. The pilot showed promise. Then it stalled because nobody owned the integration with your AMS, and the underwriters who were skeptical got their evidence: "I told you it would not stick."
You are not behind. You are exactly where 61.3% of MGAs are, per Gallagher Bassett's MGA Market Pulse. You use AI. You also do not have a formal budget for it. The 25.8 percentage-point gap between use and funding is where pilots die.
The pattern is consistent: a tool gets trialed on discretionary spend, the trial produces a real but incomplete result, leadership cannot justify a deeper investment without a clearer business case, the vendor relationship goes dormant, and the underwriting team concludes that the technology does not work. The technology is not the problem. The funding model is the problem.
Three positions a $50M operator can take
The good news: by mid-2026 the build-versus-buy decision is no longer abstract. The capital concentration makes the choice cleaner. Three positions are defensible.
Submission intake and triage. This is the lowest-friction lane. The integration is contained (email, AMS, carrier portals). The payback is measurable (submissions per underwriter per week). The vendor market is mature, with Cytora, Novella, and a handful of others all production-ready. A $50M brokerage that funds this as a line item in the next six months should expect a 30% to 50% lift in underwriter throughput within two quarters. This is the AI bet most $50M operators should make first.
Underwriting platform. A bigger commit. The Federato or Akur8 path is an 18-month integration that touches your AMS, your carrier appetite data, and your underwriter workflow at the studs. The ceiling is higher: you are not just speeding up triage, you are restructuring the underwriting function. This is the path for MGAs and brokers with ambitions to write meaningfully more business with the same headcount, or to spin up a new program or carrier appointment without a corresponding hiring round.
AI-native sub-MGA. The most aggressive position. If you have a defensible niche, balance sheet appetite, and a carrier relationship willing to capacity-back you, the Comeryx and Honeycomb model is available: stand up a fully AI-native MGA arm targeting accounts your existing operation cannot economically write. Expect $5M to $10M in build cost. Expect a 36-month payback. Expect to be one of fewer than fifty $50M operators in the country pursuing this.
Most should be on the first path within ninety days. The second is the eighteen-month decision. The third is the answer to "what does PE want when they buy us in 2028?"
The binding constraint
The vendor market is not your problem. The vendor market is mature, well-funded, and demo-ready. Your binding constraint is that you do not have a line-item AI capex budget.
AI capex inside a $50M brokerage cannot live on the credit card the same way ChatGPT Plus subscriptions do. It needs:
- A named owner with authority to make integration decisions across the AMS, the underwriting team, and the carrier relationships.
- A budget line that survives the planning cycle. Not "and AI" appended to the IT line. A standalone capex commitment, sized against a measurable throughput target.
- A 12- to 24-month roadmap with named vendors, named integration milestones, and named outcomes (submission throughput, quote-to-bind cycle time, loss ratio at year two).
- Permission for the underwriting team to redesign workflows around the tool, not just bolt the tool onto the existing process.
Carriers and the larger brokerages already have all four. That is where the McKinsey 50% capacity gains come from. The mid-market brokerage that runs three pilot vendors without a budget owns none of the four. The result is predictable.
The funding gap is the leverage point. The vendor problem is solved. The internal organization problem is not.
What the next eighteen months look like
Insurance AI spend is forecast to grow more than 25% in 2026, with the heaviest concentration in commercial-line workflows where the unit economics already support deeper investment. Gen AI adoption in underwriting is forecast to jump from roughly 14% today to 70% within three years. The PE community will continue rolling up AI-enabled MGAs and broker platforms because the EBITDA multiples (15x to 20x for tech-enabled MGAs with proprietary underwriting per Morgan Partners' 2026 insurance distribution market overview) reward operators who have made the leap.
A $50M brokerage or MGA that ends 2027 without a line-item AI budget, an owned integration roadmap, and a measurable throughput target is on a different exit path than one that does. The vendor capital is not running out. The capacity to deploy it inside an operating brokerage is the real bottleneck.
You will not catch the funded vendors by trialing more of them. You will catch them by funding the integration that turns a pilot into a workflow.
That is the conversation we have with brokerages and program administrators in the $40M to $80M band every week. We are not selling another vendor. We are helping operators turn the informal AI pilots already running in their shop into a line-item AI capex plan with named owners, measurable targets, and a four-week build cadence. If that is the conversation you need to be having in 2026, book thirty minutes with us and we will walk through what your AI line item should actually look like.
Keep Reading
- Broker Roll-Up Math at a $40M Agency: Companion strategy piece on PE roll-up economics in mid-market insurance and what that means for the AI capex conversation your board is starting to have.
- AI for Mid-Market Insurance: What Actually Works: The operator playbook on which AI workflows ship quickly in a $50M brokerage and which stall in pilot purgatory.