Risklytics launches insurance brokerage for AI companies
Risklytics, a company in Y Combinator's Summer 2026 (S26) batch, launched on Hacker News as an insurance brokerage built specifically for companies putting AI to work, from software agents handling customer requests to robot fleets operating in the field.
The company's stated thesis is that an AI agent handling customer work, or a robot on a customer site, carries risk that standard commercial insurance forms never contemplated. Carriers still quote this risk anyway, on paper drafted before AI-driven and robotic operations existed, and the exclusions buried in that paper only surface after a loss. Risklytics says it takes an applicant's operation to carriers that want this risk on their books, and reads every policy form that comes back before anything binds.
The application process has three steps. First, the customer answers three short questions, in plain English, about what they build and where it touches the physical world. Second, an AI system turns those answers into a draft insurance application, which the customer reviews and confirms field by field before it goes anywhere. Third, the submission goes to carriers willing to underwrite this kind of risk; a licensed producer reviews each submission, and nothing is bound online.
Customers pick individual coverage lines inside the application, or start from ready-made packages: one aimed at teams shipping a first product or running a limited pilot, another for AI and machines already working in production at customer sites, and options with higher limits as a company's revenue, headcount and deployments grow. Risklytics is explicit that these packages are only application starting points, not evidence of coverage; carrier appetite, policy terms, licensing and availability govern every actual quote.
Key facts
- Risklytics, a Y Combinator Summer 2026 (S26) company, launched on Hacker News as an insurance brokerage for companies deploying AI agents and robots.
- Its thesis: AI agents handling customer work or robots on customer sites carry risk that standard commercial insurance forms never contemplated, and exclusions in older paper surface only after a loss.
- The application flow has three steps: three plain-English questions from the customer, an AI-drafted application the customer reviews and confirms field by field, then carrier submission reviewed by a licensed producer, with nothing bound online.
- Packages range from coverage for teams shipping a first product or running a limited pilot up to higher-limit options for AI and machines already in production at customer sites.
- Risklytics states its packages are application starting points, not evidence of coverage, since carrier appetite, policy terms, licensing and availability govern every quote.
Why it matters
Commercial insurance forms predate agentic AI and field robotics, so risks like an AI agent acting on a customer's behalf or a robot working at a customer site fall into gaps that older policy language was never written to cover. Risklytics is positioning itself as a specialist broker that reads carrier paperwork closely enough to catch the exclusions that would otherwise surface only after a claim, and that places this risk with carriers specifically willing to underwrite it.
Who it affects
The product targets companies that put AI to work in ways that touch the physical world or handle customer-facing operations directly, from small teams shipping a first AI product or running a limited pilot to companies already running AI and machines in production at customer sites. It is aimed at founders who need commercial coverage but whose operations do not fit standard policy templates.
How to use it
A prospective customer goes to Risklytics, answers three plain-English questions about what they build and where it touches the physical world, then reviews and confirms an AI-drafted application field by field. The submission is sent to carriers that want this risk and is reviewed by a licensed producer; nothing binds online. Customers can select individual coverage lines directly or start from one of the pre-built packages before it goes to broker review.
How solid is it
The claims here come from Risklytics's own launch page and Hacker News post, not an independent source; as of the post being read, it had 50 points and 19 comments on Hacker News about 15 hours after posting. The site names no carrier partners, gives no pricing or premium figures, and states no specific launch date, so the offering's actual underwriting relationships and cost are unverified from the text alone.
Risks and caveats
No carrier names are disclosed, so it is unclear which insurers actually back these policies. No pricing or premium figures are given anywhere in the source. Beyond the YC S26 batch label, no funding amount or investor names are stated, and no founder name or team background appears in the article body. No customer names or claims case studies are provided. Risklytics itself cautions that its packages are application starting points, not evidence of coverage, and that carrier appetite, policy terms, licensing and availability govern every quote.
“An AI agent handling customer work, or a robot on a customer site, carries risk that standard commercial forms never contemplated. Carriers quote it anyway, on paper drafted before any of this existed, and the exclusions buried in that paper surface after a loss.”
— Risklytics, company launch page