When Your AI Gets It Wrong: How Business AI Tools Create New Liability

Businesses have adopted AI faster than almost any technology in memory. Chatbots answer customer questions. AI tools draft proposals, marketing copy, reports, and client communications. Software screens job applicants and summarizes documents. Most of it saves real time. 

What gets less attention is that every one of those tools speaks on your behalf. When an AI tool gives a customer wrong information, produces flawed work, or uses data it should not, the liability does not belong to the software. It belongs to you. And insurance policies are only beginning to catch up with that reality. 

The Case That Set the Tone 

In February 2024, a British Columbia tribunal decided a dispute that has become the reference point for AI liability. A traveler asked Air Canada’s website chatbot about bereavement fares and was told he could book at full price and apply for the discount afterward. That was wrong. The airline’s actual policy, published elsewhere on its own site, required the request before travel, and Air Canada denied the refund. 

The airline argued that it could not be held responsible for what its chatbot said. The tribunal disagreed. As the American Bar Association’s analysis of the decision explains, the tribunal found Air Canada liable for negligent misrepresentation, reasoning that the chatbot was part of the airline’s own website and could not be treated as a separate entity. The fact that correct information existed on another page did not excuse the error. 

The damages were small. The principle was not. If your business deploys an AI tool, what it tells people can be treated as what you told them. 

Where AI Liability Comes From 

1. Customer-facing misinformation. A chatbot that misstates pricing, policies, eligibility, or terms can create misrepresentation claims, exactly as in the Air Canada case. 
2. Errors in professional work. Professional services firms using AI to draft advice, analysis, reports, or deliverables remain responsible for the accuracy of that work. An AI-generated error in a client deliverable is a professional error. 
3. Content and intellectual property. AI-generated text and images can raise intellectual property, defamation, and advertising claims if they reproduce protected material or make false statements about others. 
4. Privacy and data handling. Entering customer or employee information into AI tools can create privacy exposure if that data is retained, disclosed, or handled in ways your obligations do not allow. 
5. Employment decisions. AI tools used in hiring, scheduling, or evaluation can produce outcomes that lead to discrimination claims, and the employer remains accountable for the decision. 

This is a different risk from AI-powered cyberattacks, where criminals use AI against you. Here the source of the problem is a tool you chose to deploy. 

Will Your Insurance Respond? 

The honest answer is that it depends on the claim and the policy, and the ground is shifting. 

1. General liability. General liability is built around bodily injury, property damage, and personal and advertising injury. Many AI-related losses, such as financial harm from bad information, fall outside those categories. 
2. Professional liability. Errors and omissions coverage is often the most relevant policy for flawed advice or deliverables, but policy wording varies and some forms are beginning to address AI specifically. 
3. Cyber liability. A cyber liability policy may respond to privacy and data incidents involving AI tools, depending on how the incident occurred and how the policy is written. 
4. Media liability. Media liability coverage can address certain content, defamation, and intellectual property claims. 

The bigger change is happening at renewal. Beginning in January 2026, Verisk, which publishes the standardized ISO forms many carriers use, made generative AI exclusion endorsements available for general liability policies, and some carriers have developed their own AI exclusion language. Carriers are not required to use them, but a policy that renewed this year may contain an AI exclusion that was not there before. The only way to know is to read the renewal. 

Practical Steps to Manage the Risk 

1. Inventory your AI use. List every AI tool in use, who uses it, and what it touches: customers, client work, employee data, or public content. 
2. Keep a human in the loop. Treat AI output as a draft, not a final answer, wherever it reaches customers or clients. The Air Canada case shows that a correct answer elsewhere on your site does not cure a wrong one from your chatbot. 
3. Control the data. Set clear rules about what customer, client, and employee information may be entered into AI tools, and which tools are approved. 
4. Review vendor contracts. Check your agreements with AI vendors for indemnification, data-use terms, and limits of liability. Many vendor terms place the risk squarely on you. 
5. Read your renewals and tell your broker. Ask whether your current or renewing policies contain AI exclusions, and tell your agent how you use AI so coverage gaps can be addressed rather than discovered. 

Review Your Coverage for the AI Era with Inszone 

AI is a genuine advantage for businesses that use it carefully. The liability that comes with it is manageable when you know where it sits. An Inszone commercial agent can review how your general liability, professional liability, cyber, and media coverage would respond to an AI-related claim, check your renewals for new exclusions, and help close the gaps. Contact Inszone Insurance to start the review. 

Frequently Asked Questions 

Is my business responsible for what its chatbot tells customers? 

Courts and tribunals have treated it that way. In the Air Canada case, the tribunal rejected the argument that a chatbot was a separate entity and held the company liable for its misinformation. 

Does general liability cover AI mistakes? 

Often not. General liability focuses on bodily injury, property damage, and personal and advertising injury, and AI exclusion endorsements became available for general liability policies in January 2026. 

What coverage is most relevant for AI errors in client work? 

For professional services firms, errors and omissions coverage is typically the most relevant, though wording varies by policy. 

Do disclaimers on my chatbot protect me? 

Disclaimers may help, but they are not a guarantee. Accuracy controls, human review, and appropriate coverage are a far stronger combination.