Silent AI
Hidden Risk, Real Exposure
Source: "Not So Silent: Tackling the Complexities of AI Liability," Gallagher, 2026
AI Risk By the Numbers
1 in 5
Insurers report AI-related losses.
Only 50%
Of those losses are fully covered
AI Adoption Is Outpacing Risk Clarity
As organizations deploy third-party AI, new risks are emerging faster than insurance frameworks can adapt.
AI adoption
Coverage clarity
Risk gap
What Is “Silent AI”?
AI-related risks that are not clearly covered or excluded in existing policies.
This creates uncertainty when losses occur.
Where Risk Shows Up
Two broad categories of AI technology systems behind the exposures:
Generative AI – large language models, chatbots and content generation platforms
Non-generative AI – such as the traditional machine learning-driven systems
Generative AI
Risk from outputs (content, decisions, recommendations)
Non-Generative AI
Risk from failures (data issues, model errors, performance breakdowns)
Two broad categories of AI technology systems behind the exposures:
Generative AI – large language models, chatbots and content generation platforms
Non-generative AI – such as the traditional machine learning-driven systems
Where Risk Shows Up
Who's Responsible?
Risk is hard to define—and even harder to assign
Data provider
Data provider
Developer
Builds the AI
Deployer
Integrates AI
End user
Operates AI
