AI Risks Expose Gaps in Cyber Insurance Coverage

Cyber insurance underwriting relies heavily on applications, attachments and email exchanges that often arrive in inconsistent, unstructured formats. That fragmentation can leave submissions trapped in a review “black hole,” slowing risk assessment and policy decisions. Insurance technology provider IntellectAI is applying Agentic AI and large language models to organize the material and identify risk attributes more efficiently for underwriters.

IntellectAI has highlighted Magic Submission, an intelligent document-processing tool designed to extract cyber-risk information from unstructured files and emails and convert it into data usable in underwriting workflows. The company says the system can improve submission handling, review efficiency and assessment accuracy. The report did not disclose a launch date, customer count, measured reduction in processing time, deployment cost or investment amount.

Companies racing to deploy generative artificial intelligence face emerging liabilities that conventional commercial policies often exclude. San Francisco-based insurance startup Corgi has introduced coverage designed specifically for AI systems, targeting risks including algorithmic bias, hallucinated or inaccurate content, and disputes over training data. The product aims to help businesses transfer potential legal, compensation and operational costs arising from AI deployment.

As of August 2026, Corgi said the policy covers six major categories of AI-related failures, filling gaps left by traditional insurance products. The company has raised $108 million and surpassed $40 million in annual recurring revenue, or ARR. Its rapid fundraising and revenue growth underscore rising corporate demand for specialized protection as businesses put AI models into production faster than existing risk-management frameworks can adapt.