AI is entering a new phase. Instead of simply generating content or answering questions, AI agents are increasingly being trusted to make decisions, interact with systems and complete tasks on behalf of users.
The business case is clear – models are capable of transforming efficiency, driving lower costs and higher output. But while autonomous AI does generate significant opportunity, it also introduces a new set of risks that don't fit neatly into traditional cyber or technology exposures.
How does autonomous AI work, and why does it matter for insurance?
Autonomous AI differs from traditional software because it isn't limited to following a fixed set of instructions. Given a goal, it can gather information, make decisions and take action with limited human involvement. This makes it harder to predict exactly how the technology will behave, and harder to place responsibility when anything goes wrong.
For the insurance industry, this puts accountability and liability into the spotlight. If an autonomous AI system causes harm without malicious intent, who is responsible? As AI adoption accelerates, it's creating risks that sit somewhere between technology, cyber and professional liability, challenging traditional approaches to insurance.
Risks for AI specialists
Third-party risk: While traditional professional liability claims stem from services provided to customers, autonomous AI agents can create liabilities involving unrelated third parties. If an AI agent acts unexpectedly and causes harm outside the intended use case, accountability questions can quickly emerge.
Unauthorized access: Incidents involving unauthorized access, data exposure or operational disruption may create claims even where there was no malicious intent behind the AI's actions.
Product liability: As AI systems become more autonomous, liability may not always depend on proving a software bug or developer error. Emerging legal frameworks are increasingly exploring whether AI-enabled products should attract product liability exposure in the same way as other products brought to market.
Blurring responsibilities: As AI systems perform tasks traditionally carried out by humans, the distinction between technology product risk and professional service risk becomes less clear.
Risks for AI users
Growing AI reliance: AI adoption is accelerating rapidly, with 88% of businesses now using AI in at least one function, up from 78% in the previous year. As more businesses embed AI, understanding the risks is becoming vital.
Error propagation: If an AI-enabled workflow, recommendation engine or autonomous agent produces an error, that mistake can flow directly into client advice, reports, recommendations and decision-making.
Professional responsibility: AI users remain responsible for the outcomes of their AI systems. They cannot assume liability can be passed to the AI creator and are accountable for deploying and managing those systems safely.
AI-built applications: The rise of AI-assisted software development and "vibe coding" is enabling organizations without traditional software engineering expertise to build increasingly sophisticated applications and workflows.
Accountability: Organizations need a clearer understanding of where responsibility sits when AI plays a role in delivering services, making decisions or interacting with customers on their behalf.
Affirmative coverage: Keeping pace with AI risk
With the impact of autonomous AI set to grow, businesses need confidence that their insurance keeps pace with how these technologies work.
To help protect your clients, look for:
Affirmative AI coverage that reduces uncertainty around AI-related exposures.
Specialist underwriting expertise to understand evolving AI risks.
Proactive risk prevention to help organizations deploy AI safely and responsibly.
Clearer guidance as legal, regulatory and liability frameworks continue to evolve.
Clear, affirmative AI wording is key to providing certainty around how AI-related exposures will be treated. As AI adoption accelerates, it's time to give businesses greater clarity, reduce ambiguity when claims arise and strengthen resilience.
Get in touch to learn more about AI risks and how CFC will be providing affirmative AI coverage.
