As AI incidents, regulation, and litigation increase, corporate buyers are seeking dedicated insurance for AI solutions, from classic ML to generative AI applications to AI agents. The Geneva Association found that nine in ten of 600 corporate insurance buyers surveyed signaled a need for dedicated coverage for generative AI (Gen AI Risks for Businesses, October 2025). Armilla's liability coverage responds to AI risks directly, giving clients clarity as they scale AI.

Armilla AI is a Lloyds backed managing general agent (MGA) and recognized pioneer of underwriting and insurance solutions for artificial intelligence.
Backed by A-rated global insurers including Chaucer, Axis Capital, and Convex, Armilla offers affirmative AI liability insurance empowering companies to deploy AI with confidence.


Superior
AM Best


Excellent
AM Best


Excellent
AM Best
Pre-AI Policies Are Silent on AI
If you sell or deploy AI to large enterprises, proof of insurance covering AI risk is increasingly a condition of the deal — written into MSAs, SLAs, and procurement reviews. The tech E&O and cyber package most companies carry was designed before generative AI, which makes it silent on AI: it neither clearly affirms coverage for AI failure modes nor clearly denies it. Silence is hard to hand a procurement team as proof.
Two points are worth stating plainly. The contractual liability an AI indemnification clause creates typically sits behind a contractual-liability exclusion in tech E&O today. And insurers are actively exploring AI exclusions across traditional lines; most are not in force yet, but the reviews are underway, on faster timelines than silent cyber. Affirmative, named AI coverage answers the question procurement is actually asking, rather than relying on what an older policy might pay.
Standalone, affirmative coverage that responds to AI risks.
Sample Coverage Areas
AI Model Error Liability: Financial loss suffered by third parties arising from AI model underperformance (e.g. errors, inaccurate outputs, hallucinations).
AI Model Output Liability: Legal claims arising from the outputs of the underperforming AI system, including breach of confidentiality, trade secrets, defamation, libel, and slander.
AI Agent Mistakes: Legal claims arising from underperforming AI agents, including failure to escalate recommendations, incorrect decisions causing damages.
Non-Breach Privacy & Data Leakage Liability: Legal claims and regulatory investigations arising from unintended disclosures through model outputs, hallucinations.
AI Model Property Damage Liability: Damage to third-party property caused by the underperformance of an AI system.
AI Model Regulatory Violations: Defense costs and fines for actions or investigations into AI underperformance, including “high-risk” AI systems (e.g. Colorado AI Act, EU AI Act).
Through our partnership with Lloyd's of London and our work with enterprise AI leaders, we develop coverage at the pace of AI adoption.
Regulated Services
Financial Services, Healthcare, Telecommunications & Media
✓ COVERED
Professional Services
Legal Services, Public Affairs, Customer Services, Call Centres
✓ COVERED
Industrial & Infrustructure
Real Estate, Construction, Manufacturing
✓ COVERED
Enterprise Technology
CRM, Supply Chain, Data Unification, and more
✓ COVERED