Shoppers who spot AI in your store trust it less.
AI writes product descriptions, makes lifestyle images, answers customer questions and summarises reviews. It's quick and cheap. But shoppers are getting good at spotting it, regulators are watching reviews closely, and courts have made it clear that what your chatbot promises, you deliver.
The rules and the numbers.
Shoppers notice, and buy less.
53% of consumers say they're less likely to buy from brands they know use AI in their social media content.
Fake reviews now carry fines.
The Federal Trade Commission's rule on consumer reviews bans fake reviews, including AI-written ones, with penalties of up to $53,088 per violation. In December 2025 the FTC sent its first warning letters under the rule to nearly a dozen companies.
Your chatbot speaks for you.
Several states, including California, Utah, Maine and New Jersey, require chatbots to say they're AI in some situations.
Source: Cooley on state chatbot laws
Trackers need a choice.
24 states now have consumer privacy laws that give shoppers rights over the data your store collects, including through plugins and pixels.
Source: MultiState, October 2026
Real cases.
Air Canada, 2024
The airline's website chatbot told a grieving customer he could claim a bereavement discount after travelling. The policy said otherwise. Air Canada argued the chatbot was responsible for its own words. The tribunal called that a "remarkable submission" and made the airline honour what the chatbot said.
Your chatbot's promises are your promises.
Source: Pinsent Masons, February 2024
Priceline, 2025
Priceline's chatbot, Penny, told a customer their non-refundable hotel booking qualified for a full refund. They cancelled. The refund never came, and it took almost six months, and a national newspaper's consumer column, before Priceline admitted the chatbot had given "misinformation" and paid up.
A chatbot that guesses costs you money and goodwill, and the story gets told.
Source: The Telegraph consumer column, May 2026, as reported by AI Vibe Graveyard
Where it goes wrong, and how we fix it.
Product descriptions that read as AI.
Generic phrases ("elevate your everyday", "seamless", "crafted for you") tell shoppers nobody who knows the product wrote it.
Starter reads your pages every week, shows which ones read as AI and why, and suggests a rewrite in your own voice. Need it done for you? Our Edit add-on rewrites the page, and a person checks it.
Images that look AI-made.
Waxy skin, odd hands and too-perfect rooms make a product look less real.
Starter checks your images every week and flags the ones customers are likely to see as AI. Our Design add-on finds real stock photos to replace them, and a person checks the picks.
Reviews you can't stand behind.
AI-written reviews, reviews from staff, or only the best reviews shown.
Starter flags reviews and testimonials that look invented. Plus gives your team clear rules on reviews, so nobody writes one "to get things started".
A chatbot that doesn't say it's AI, or makes things up.
Starter spots AI chat tools on your store and asks you to confirm your chat tells shoppers it's AI. Plus sets what your chatbot may and may not say (refunds, prices, delivery dates) and who checks it.
Plugins quietly collecting customer data.
Pixels, review apps and chat tools can load before anyone agrees, and some use your customers' data to train their AI.
Starter lists every tool on your store, shows what its terms say about AI and your data, and flags trackers that load before shoppers choose.
What it looks at.
Your homepage and a few key pages, the images on them, the tools your store loads, your privacy notice, and whether your chat says it's AI. You get a score, your top three gaps and a plan recommendation in about 30 seconds.
See how your store reads to shoppers.
Your website, a few questions, and the three things to fix first.