Insurance Industry Splits on AI Agent Coverage — Some Insurers Add "Absolute AI Exclusion" Clauses While Others Launch Agent-Specific Policies

A major AFP wire story published March 15, 2026, syndicated across Reuters, Economic Times, Digital Journal, RTL, and dozens of outlets, reveals a deepening split in the insurance industry over how to handle AI agent liability.

THE DIVIDE:

As businesses increasingly deploy agentic AI systems to autonomously grow revenues, make decisions, and interact with customers, the insurance industry is fracturing into two camps:

1. EXCLUSION CAMP: Some standard insurance policies now include "absolute AI exclusion" clauses that expressly deny coverage for any AI-related mishaps. These insurers view AI agent behavior as too unpredictable and the risk too difficult to model.

2. COVERAGE CAMP: Specialized insurers are stepping in to fill the gap. Companies like Armilla AI now offer policies specifically covering AI agent errors, hallucinations, and malfunctions — creating an entirely new insurance product category.

REAL-WORLD EXAMPLES:

• A commercial real estate firm attempted to have its AI agent covered as a regular employee under its existing business insurance. The insurer rejected this and required a separate special policy. This illustrates the fundamental problem: existing insurance frameworks were not designed for autonomous digital workers.

• The incident mirrors broader trends where AI agents make consequential errors — from the Amazon Kiro agent that deleted an entire AWS Cost Explorer environment (13-hour outage) to banking agents attempting $5,000 refunds for $500 disputes.

WHY THIS MATTERS FOR AI AGENT DEPLOYMENT:

Insurance coverage (or lack thereof) is becoming a gating factor for enterprise AI agent adoption. Companies deploying agents for customer-facing tasks, financial operations, or operational decisions need clarity on liability. The current uncertainty means:

• Enterprises may delay agent deployment until insurance frameworks mature
• Companies using agents may face uncovered losses from agent errors
• A new market for AI agent insurance is emerging, estimated to become significant as the agentic AI market grows toward the projected $139B by 2030 (Morgan Stanley)

REGULATORY IMPLICATIONS:

The insurance industry split signals that even the financial sector cannot yet agree on how to classify AI agent risk. This will likely accelerate regulatory frameworks for AI agent liability, particularly in the EU where AI Act implementation is underway, and in sectors like finance and healthcare where autonomous decision-making carries high stakes.