Mount Review 2026: The AI Agent Insurance Carrier
Mount is a Y Combinator-backed, AI-native insurance carrier built specifically for companies deploying autonomous AI agents. Where a traditional carrier prices coverage off a form, Mount tests the agent itself, then insures what is left. This is an independent profile from QuoteSweep, which maps the emerging AI agent insurance category. QuoteSweep does not compete with Mount.
TL;DR: Mount (mount.insure) is an AI-native insurance carrier for deployed AI agents. Its product, ActiveCover, combines continuous vulnerability scanning and red-teaming with liability coverage across eight AI-specific risk categories, and advertises 48-hour activation and dynamic pricing. It is Y Combinator-backed (Spring 2026), founded by Fabian Amherd and John Bachmann.
What Mount is
Mount describes itself as providing vulnerability identification and insurance for AI agents, combining "real-time vulnerability scanning and risk scoring" with coverage designed for agentic systems. It positions itself as an AI-native carrier: a highly automated version of a traditional commercial insurer, where underwriting, pricing, issuance, and claims are automated rather than run through manual workflows and layers of intermediaries.
How it works
Per its site, Mount operates in three integrated steps:
- Test your risk — continuous scanning identifies threats in deployed agents before attackers find them
- Improve security — automated remediation suggestions with compliance analysis and ongoing hardening
- Active insurance — coverage with 48-hour activation, dynamic pricing, and AI-specific terms
The model treats the security assessment and the insurance as one product: Mount scans and red-teams a deployment, quantifies the operational risk, helps fix vulnerabilities, and insures the residual risk.
What ActiveCover covers
Mount's product, ActiveCover, addresses eight AI-specific risk categories, including:
- Data restoration
- Business interruption
- Incident response
- IP infringement claims
- Infrastructure cost overages
- Payment errors
- Data breaches
- Service failures
The site does not publish coverage limits or premiums.
Who it is for
Mount points at AI-native companies and those running production AI deployments that need enterprise-grade protection. Its stated focus is mid-market and enterprise companies putting autonomous agents into real business processes.
Company background
- Backing: Y Combinator (Spring 2026 batch), launched May 2026
- Founders (per reporting): Fabian Amherd, who studied CS/ML at ETH Zürich, and John Bachmann
- Positioning: describes its product as purpose-built AI liability coverage for autonomous agents
How Mount compares
- vs. Klaimee and AIUC: all three pair coverage with agent assessment, but Mount emphasizes continuous, real-time scanning and remediation ("active" insurance) rather than a point-in-time certification.
- vs. Corgi: Corgi is a broad startup insurance carrier where AI coverage is one module; Mount is built specifically around insuring deployed agents.
Frequently Asked Questions
What does Mount insure?
Deployed autonomous AI agents, through its ActiveCover product, across eight AI-specific risk categories including data restoration, business interruption, IP infringement, payment errors, data breaches, and service failures.
How is Mount different from traditional insurance?
Mount is an AI-native carrier that pairs continuous vulnerability scanning and red-teaming of the agent with the coverage itself, and automates underwriting and claims rather than running manual workflows.
How fast can Mount provide coverage?
Its site advertises 48-hour activation along with dynamic pricing and AI-specific coverage terms.
Who backs Mount?
Per its Y Combinator page and reporting, Mount is Y Combinator-backed (Spring 2026) and was founded by Fabian Amherd and John Bachmann.
Explore coverage with Mount
If you deploy autonomous AI agents in production, Mount is one of the new options to compare.
For the full category, see AI agent insurance explained.
Sources: mount.insure (product, model, risk categories) and Mount's Y Combinator page and reporting (backing, founders, launch). Last verified July 7, 2026.