What is Covered?
We insure organisations that use AI in products and workflows. Our coverage is designed around the specific risks of deploying AI, giving policyholders the most relevant protection possible.
Data/Model Restoration & Retraining
Costs to restore corrupted data and retrain AI models after an incident.
Business Interruption
Revenue loss from AI-driven failures that disrupt operations.
Incident Response & Regulatory Costs
Forensic investigation, legal counsel, and regulatory notifications.
Runaway Usage & Infrastructure Overage
Unexpected spikes in API and infrastructure costs during security events.
Incorrect Funds Transfer
Financial losses from AI-driven errors in payment processing.
Illustrative Scenarios
Real-world examples of when coverage applies.
Third-Party: Harmful or Infringing AI Outputs
Misrepresentation & Misselling
Your AI assistant overstates product capabilities to a prospect, leading to a misselling claim when the product underdelivers — or generates misleading comparisons that prompt a competitor to pursue legal action for misrepresentation.
First-Party: Business Interruption
Costly Operational Errors
An AI agent acting on faulty logic greenlights a bulk purchase order, misroutes inventory across warehouses, or schedules crews to the wrong site — each mistake cascading into wasted resources, contractual penalties, and costly manual corrections.
First-Party: Runaway Usage & Infrastructure Overage
Infrastructure Overage Attack
A botnet floods your public endpoint with complex, long-context queries during a security event, spiking LLM/API usage and forcing emergency mitigation spend.
How it works
Insurance built for the AI era.
Agent-by-agent risk assessment
We assess each AI agent based on its specific risk exposure — deployment context, permissions, data access, usage volume, and the users interacting with it — and price coverage accordingly. Our algorithmic underwriting engine enables rapid, consistent policy decisions, delivering quotes in minutes rather than weeks without sacrificing analytical rigor.
Learn more