Jack Kudale, Cowbell's founder and chief executive, put a number on the shift on July 28, 2026: OMNI, the AI-native decision intelligence system now running beneath the carrier's excess and surplus cyber book, compressed product deployment from roughly eight months to six weeks and contributed to a 53% increase in new business (Cowbell, July 2026).

A carrier that can ship a product in six weeks can ship a mispriced one in six weeks.

Key Takeaways

  • The deployment cycle fell from eight months to six weeks, which removes the runway an actuarial team previously had between a product concept and the first bound policy.
  • Nonadmitted business is largely exempt from prior-approval rate and form filing, so no state rate analyst reviews an E&S cyber product before it binds. The internal control is the only brake.
  • Small and midsize enterprises are roughly 98% of cyber claims by volume but about 49% of total incident cost, at a five-year average of $246,000 per SME incident against roughly $10.3 million for large enterprises.
  • Surplus lines carriers posted an incurred loss ratio near 56 in 2025 against roughly 50.2 for admitted carriers, on flat total cyber premium of $7.5 billion.
  • The pricing signal itself is contested: eight consecutive quarters of decline through Q1 2026 against an S&P Global Ratings expectation of 15% to 20% increases in 2026.

What OMNI Runs Underneath the Book

Cowbell describes OMNI as combining decision intelligence, orchestration and governance into a single AI-native operating model rather than a bolt-on assistant. Specialized bi-directional AI agents and small language models gather submission information, analyze risk signals, assess appetite and orchestrate workflows across intake, claims and cyber services, producing what the company calls decision-ready recommendations at each step. A governance layer named Bellwether provides observability into how the agents perform.

Co-founder and chief product officer Rajeev Gupta drew the line against the assistant tools that dominated carrier AI announcements through 2025: "Insurance doesn't need another chatbot." The distinction is real for governance purposes. A copilot an underwriter can ignore is a different object from a system that orchestrates decisions across the lifecycle.

Cowbell is explicit that human underwriters retain final decision authority, which keeps legal responsibility for a bound policy with the underwriter and the actuarial function rather than the agent that drafted the recommendation. The first major application targets non-admitted quotes for small and midsize organizations, against a risk pool the company says spans more than 55 million entities worldwide, with quotes that once took days delivered in minutes. That speed is what wins broker flow, and it is the mechanism behind the 53%.

The Filing Gap That Makes Six Weeks Possible

The reason a release cycle can compress this far without tripping a regulatory gate is structural rather than technological. Surplus lines carriers write on a nonadmitted basis, and nonadmitted property and casualty business is largely exempt from the prior-approval rate and form filing that governs admitted markets (NAIC).

An admitted auto or homeowners carrier proposing a 53% volume shift in six weeks would file rate pages and an actuarial memorandum before launch. An E&S carrier faces no equivalent external checkpoint on rate adequacy before it binds. The loss-cost trend, the increased limits factor curve behind an excess layer, and the comparison of the model's learned appetite against actual bound experience all still have to exist. What changed is that they now have six weeks of runway instead of eight months.

The book this runs on is where that matters most. Small and midsize enterprises account for roughly 98% of cyber claims by volume but only about 49% of total incident cost, at a five-year average incident cost of $246,000 against roughly $10.3 million for large enterprises, with SME severity up nearly 30% year over year (NetDiligence, 2025, across more than 10,000 claims from 2020 through 2024).

High frequency at low severity is the mix that should build credibility fastest, but only if the business mix holds still long enough for triangles to mature. A book growing new business 53% is diluting its seasoned experience with undeveloped exposure while the agents are still learning what to write.

The market signal the pricing rests on is itself unsettled. Surplus lines carriers now write nearly two-thirds of US cyber premium, and their incurred loss ratio reached almost 56 in 2025 against roughly 50.2 for admitted carriers, on total cyber premium essentially flat at $7.5 billion and eight consecutive quarters of pricing declines through the first quarter of 2026 (Insurance Journal, July 2026).

Third-party claims, the slower-developing liability side, grew roughly 30%. "As long as pricing continues to decline, insurers will have difficulty reversing the increasing loss ratio," AM Best's analysts wrote. S&P Global Ratings expects the opposite, premium increases of 15% to 20% in 2026 on severity and AI-driven attack costs.

Observability the Carrier Owns Is Not a Record a Regulator Can Examine

Roughly 25 states have adopted the NAIC's AI model bulletin, requiring a written AI systems program covering governance, risk management and internal controls over predictive models, and the NAIC's Big Data and Artificial Intelligence Working Group has an AI Systems Evaluation Tool piloting across a dozen states with broader adoption anticipated at the 2026 Fall National Meeting (NAIC).

Colorado went further on mechanics, building a quantitative testing regime for algorithms and predictive models under SB21-169 and expanding the governance regulation in 2025 beyond life to private passenger auto and health plans using external consumer data. Cyber E&S is not in that scope.

SystemArchitectureReported outcomeGovernance layer
Cowbell OMNIProprietary, end-to-end, built in-house53% new-business growth; 8-month to 6-week deploymentBellwether (internal, carrier-controlled)
Sixfold AI UnderwriterWalled, carrier-specific fine-tuned layer15%+ hit-ratio gain; up to 30% GWP per underwriterCarrier-run validation against ASOP No. 56
Duck Creek Agentic AIFive-layer orchestration, separately priced modulesIncremental adoption across underwriting and claimsVendor-supplied, module-level

Bellwether points in the same direction those frameworks do, and differs in who can inspect it. It is a proprietary tool built and controlled by the carrier whose volume it enables, where the NAIC and Colorado regimes are external requirements a regulator examines. An internal observability layer is a necessary first step, not a substitute for a documented, auditable validation record.

The build choice compounds that. Sixfold's AI Underwriter launched in June 2026 across six carriers representing $270 billion of gross written premium, reporting hit-ratio gains of 15% or more, and sells a walled carrier-specific layer that plugs into an existing book (The Insurer, June 2026), as covered at launch. Duck Creek sells separately priced orchestration modules a carrier can adopt incrementally, one of several vendor architectures competing for the same layer.

Building native has a genuine advantage: instrumentation can capture a decision at the point it is made rather than reconstructing an audit trail afterward. The cost is that Cowbell owns the entire validation burden, with no vendor sharing the model-risk testing or the regulatory defense of the pricing logic beneath a book that is more than half again larger in new business than it was.

Further Reading

Sources