A Client Just Asked About Your AI Use in a Contract. Are You Covered?

By Brent Dixon, Licensed P&C Commercial Insurance Advisor and AI Risk Consultant · B. Dixon Risk Management Corp. · bdixonrisk.com

It used to be a single line in a vendor questionnaire. Now it is a full section of the contract. Clients want to know whether you use artificial intelligence, what data you put into it, who reviews the output, and who pays if something goes wrong. Some are asking you to warrant that you will not use AI at all without permission. Others want you to indemnify them for any AI-related error.

Most businesses already use AI somewhere: drafting proposals, summarizing documents, writing code, answering customers, analyzing data. Very few have checked whether their insurance responds when that use leads to a claim. This guide explains what clients are asking for, which policies may or may not cover AI-related losses, and how to put your business in a position to sign with confidence.

Key Takeaways

• AI clauses are becoming standard in professional services, technology, marketing, healthcare-adjacent and financial contracts.
• No single policy covers all AI risk. Professional liability, cyber, media and general liability each respond to different pieces.
• Exclusions and application questions are evolving. Some carriers are adding AI-specific questions and exclusions, so last year’s coverage review may already be outdated.
• Contract promises can exceed your coverage. Broad warranties and uncapped indemnities can leave you personally exposed.
• Written AI governance helps with clients, with underwriters, and with defending a claim.

What Clients Are Actually Asking

Clients are not anti-AI. They are trying to manage their own risk: confidential data leaking into public tools, inaccurate output reaching their customers, intellectual property problems, and regulators asking how decisions were made. Their contracts and questionnaires reflect those concerns.

Each of these is reasonable on its own. The problem is that many businesses sign them without checking whether their actual practices match, or whether their insurance would back up the promises if a client made a claim.

Which Policy Responds to an AI Claim?

AI does not create a brand-new category of loss so much as a new way to cause familiar ones: a professional error, a data breach, a copyright dispute, a discrimination allegation. The question is which of your policies covers the familiar loss when AI was involved, and whether anything in the policy carves it out.

Professional Liability / Technology E&O

If an AI-assisted deliverable contains an error that costs your client money, such as a flawed analysis, an incorrect recommendation or buggy code, this is usually the policy that responds. Coverage depends on how your professional services are defined. If the definition does not reflect the way you now deliver work, there may be room for a dispute.

Cyber Liability

If confidential client data is exposed through an AI tool, whether pasted into a public model or leaked through a compromised integration, cyber insurance is the likely responder. Look closely at how the policy treats data shared with third-party vendors and whether it requires specific security controls as a condition of coverage.

Media Liability and General Liability

AI-generated marketing content, images or copy can trigger claims of copyright infringement, defamation or misappropriation. Media liability is designed for these risks. General liability policies offer limited protection here and frequently exclude intellectual property claims, so relying on them alone is risky.

The Contract Promises That Outrun Your Insurance

Insurance generally covers your legal liability for negligence. It is much less reliable when you voluntarily take on obligations beyond that, such as guaranteeing that deliverables are error-free, warranting that no AI was used, or agreeing to indemnify the client for any loss related to AI regardless of fault. Many professional liability policies limit coverage for liability you assume under contract unless you would have been liable anyway.

That makes negotiation important. Reasonable middle ground often includes limiting indemnity to losses caused by your negligence, capping liability at a multiple of fees or at your insurance limits, replacing “no AI” warranties with disclosure and approval processes, and defining which tools are approved for client data. Your attorney should negotiate the language. Your broker should confirm that the result matches your coverage.

Build an AI Governance Starter Kit

The best way to answer a client questionnaire, and an insurance application, is with a short, written policy you actually follow. It does not need to be complicated.

Documented governance does three things at once. It lets you answer client questionnaires accurately and quickly. It gives underwriters evidence that you manage the risk, which matters as more applications ask about AI. And if a claim happens, it shows that your business acted reasonably, which is the foundation of most professional liability defenses.

Frequently Asked Questions

Do I need a separate “AI insurance” policy?

For most small and midsize businesses, the starting point is making sure existing professional, cyber and media coverage is written broadly enough to include AI-assisted work and has no problematic exclusions. Specialized AI products exist and are developing, and they may make sense for companies that build or sell AI systems.

Should I disclose AI use on my insurance application?

Answer every application question accurately. Inaccurate answers can jeopardize coverage. If an application does not ask, it is still worth discussing your AI use with your broker so the policy is structured to match.

What if I don’t use AI at all?

Check carefully. AI is now built into email, word processing, design, accounting and customer service software. Many businesses use it without realizing it, which matters if a contract warrants that you don’t.

Sign AI Clauses With Confidence

B. Dixon Risk Management combines commercial insurance expertise with AI risk consulting. We help businesses review AI contract language, build practical governance policies, and align their professional, cyber and media coverage with the way they actually work.

Got a contract or vendor questionnaire asking about AI? Send it to us along with your current policies. We will identify the gaps between what you are promising and what you are covered for. Email bdixon@dixoninsure.com or call 786-804-2580.

This article is general information, not legal advice or a coverage opinion. AI-related policy terms and exclusions are evolving quickly and vary by carrier. Have an attorney review contract language and a licensed professional review your coverage.