Every week brings a new pitch for AI tools that promise to run parts of a business automatically: chatbots that handle customer service, systems that write marketing copy, agents that place orders and manage inventory without a person watching. The pitch is speed and cost savings. What rarely gets mentioned is who pays when the AI gets it wrong.
The clearest answer so far comes from a small claims case in Canada. In 2024, a man asked Air Canada's website chatbot about bereavement fares after a family death. The chatbot gave him wrong information, and he booked a full-price ticket based on that bad advice.
When he asked for a refund, Air Canada refused. The airline's defense in front of the tribunal was that the chatbot was responsible for its own words. The tribunal did not accept that argument, and Air Canada had to pay.
The compensation was small, under a thousand dollars. But the ruling matters far more than the dollar figure. It confirmed something every business owner needs to internalize: a company cannot point at its own software and say the mistake was not its fault.
If the tool speaks for the business, the business owns what it says. This is not staying confined to one court case.
In the United States, several states passed laws in 2026 specifically targeting chatbots, requiring disclosures when customers are talking to AI and creating new grounds for lawsuits when chatbots cause harm. Australia is rolling out its own ban on unfair trading practices, aimed partly at manipulative sales tactics, including ones powered by AI that flatter or pressure a customer into a purchase.
The direction is the same everywhere: regulators are not creating a special lighter rulebook for AI. They are applying the same consumer protection rules that already existed, and adding new ones aimed squarely at automated decision-making.
The part most business owners are missing is insurance. Standard business insurance, the kind covering general liability or professional errors, was written before AI chatbots existed. It often does not cover the specific ways AI fails, such as a chatbot confidently inventing false information.
Insurers are starting to offer separate AI coverage, but plenty of small and mid-sized businesses do not know they need it until a claim gets denied. The practical fix is not complicated, but it does require attention most business owners are not currently giving it.
Read the contract with any AI vendor and understand who is responsible when the tool fails, not just when it works. Keep a real person checking what the AI produces before it reaches a customer, especially anything involving prices, discounts, or promises.
Check your insurance policy for what it actually says about AI-related errors. The gap between what businesses assume is covered and what actually is covered is where the real risk sits.