Many small businesses now use AI somewhere on their website. It might draft blog posts, polish product descriptions, generate illustrations, translate pages, power a chat window or summarize reviews. The question that follows is simple to ask and harder to answer: should you tell visitors, and if so, how?

There is no single rule that covers every case, and the legal picture varies by country, state and industry. But a set of practical norms has emerged from search guidance, consumer protection rules and public policy work. This guide explains those norms, where disclosure clearly matters, where it is mostly a matter of good judgment, and how to write disclosures that are honest without cluttering every page.

Start with what readers would reasonably expect

The most useful test for AI disclosure is not technical. It is whether a reasonable visitor would want to know, and whether not knowing could mislead them.

Google offered a version of this test in its February 2023 post, Google Search’s guidance about AI-generated content. In an FAQ, it said AI or automation disclosures are useful for content where someone might think “How was this created?”, and suggested adding them when it would be reasonably expected. The same post said Google’s systems focus on the quality of content rather than how it is produced, and that using automation, including AI, to generate content with the primary purpose of manipulating search rankings violates its spam policies.

That framing is helpful beyond search. A recipe page written with help from an AI tool and then tested and edited by your staff probably does not need a prominent label. An image that looks like a real photograph of your shop but was generated by AI is different, because a visitor could reasonably believe it shows something real.

Where disclosure matters most

Some uses of AI carry a higher risk of misleading people. These deserve clear, visible disclosure, or a rethink.

Images that look like real photos. If a generated image could be mistaken for your actual premises, staff, products or past work, label it. Better still, use real photos for anything customers rely on to judge what they will get. Illustrative images are fine, but captions should make their nature clear.

Chat and automated conversations. When visitors type questions into a chat window, they should know whether they are talking to a person or an automated system. A short line at the start of the conversation, and an easy way to reach a human, covers most situations.

Reviews and testimonials. This is where the line is firmest, and it is covered in the next section.

Expert or regulated advice. Content about health, money, legal matters or safety calls for extra care. If AI helped produce it, the disclosure matters less than making sure a qualified person reviewed it and stands behind it.

Reviews and testimonials: a firm line

Reviews are evidence customers use to make decisions, so faking them is treated seriously. In August 2024, the Federal Trade Commission published its Trade Regulation Rule on the Use of Consumer Reviews and Testimonials in the Federal Register, effective October 21, 2024.

The rule, among other things, prohibits selling or purchasing fake consumer reviews or testimonials, and makes it an unfair or deceptive practice for a business to write, create or sell a review or testimonial that materially misrepresents that the reviewer exists, that they had experience with the product or service, or what that experience was. The rulemaking record specifically noted reporting on generative AI tools that make it easier for bad actors to write fake reviews.

The practical point is that the rule focuses on whether a review is genuine, not on what tool produced it. A testimonial attributed to a customer who does not exist is not made acceptable by a disclosure that AI wrote it. For a small business, the safe path is straightforward:

  • Publish only reviews from real customers, in their own words.
  • Do not use AI to invent, inflate or rewrite reviews in ways that change their meaning.
  • If you summarize reviews with AI, make clear it is a summary and link to the originals.
  • Disclose relationships when staff, relatives or paid partners leave reviews.

This is general information, not legal advice. Federal rules can change and states have their own consumer protection laws, so check current requirements with a qualified adviser.

Hands in peach sleeves sort printed photos of pastries and dried flowers on a purple desk beside a dark green mug.

Why disclosure builds trust

Beyond rules, there is a broader expectation forming around AI transparency. In March 2024, the National Telecommunications and Information Administration published its Artificial Intelligence Accountability Policy Report, a policy document making recommendations to government. It states that those impacted by an AI system should know when AI is being used, and notes that standardized, accessible formats and plain language can make disclosures easier to compare and understand.

The report is aimed at policymakers and AI developers, not small businesses, and it does not create obligations for them. But it reflects a direction of travel. Customers are increasingly aware of AI, and many will judge a business partly on whether it is candid about how it uses it.

The same report observes that watermarking approaches are more mature for video and photos than for text. In plain terms, technical labels embedded in content are not yet a reliable way to signal AI use across the board, which is another reason clear, human-readable disclosure on the page itself remains important.

How to write a good disclosure

Effective disclosures are short, specific and placed where they are relevant. A few patterns work well:

  • Near the content. Put a brief note at the top or bottom of an article, or in an image caption, rather than only in a policy page nobody reads.
  • Specific about the role. “Drafted with the help of an AI tool and reviewed by our team” is more informative than “AI was used.”
  • Clear about human responsibility. Say who checked the content, especially for advice customers might act on.
  • Plain language. Avoid jargon and legalistic phrasing.
  • Consistent. Use the same wording and placement across the site so visitors learn what to expect.

Good copywriting makes disclosures read as part of a trustworthy voice rather than a nervous disclaimer.

Example wording you can adapt

Disclosures do not need to be elaborate. These examples show the level of detail that usually works, and can be adjusted to fit how your business actually operates:

  • For an article: “This guide was drafted with help from an AI writing tool, then fact-checked and edited by our service manager.”
  • For an image caption: “Illustration created with AI. Not a photo of our shop or staff.”
  • For a chat window: “You’re chatting with our automated assistant. It can answer common questions. Type ‘person’ at any time to reach our team during business hours.”
  • For a review summary: “This summary was generated automatically from customer reviews. Read the full reviews below.”

Each example says what AI did, what it did not do, and where a person is involved. Only use wording that is true for your setup. A disclosure that overstates human review is its own kind of misleading claim.

Bylines and authorship

Google’s 2023 guidance also addressed bylines. It suggested considering accurate author bylines where readers would reasonably expect them, such as content where someone might ask, “Who wrote this?” It added that giving AI an author byline is probably not the best way to make clear to readers when AI is part of the content creation process.

That points toward a sensible norm: credit the person who is responsible for the content, and describe AI’s role in a note rather than presenting a tool as the author.

A woman in a green coat and peach scarf reads her phone at a bus shelter with a purple tote on her shoulder.

Design disclosures into the site

Disclosure works better when it is planned into the design rather than bolted on. Consider:

  • A consistent caption style for illustrative or generated images.
  • A clear label at the top of any chat window, with a visible link to contact a person.
  • A short page explaining how the business uses AI, linked from the footer, covering content, images, chat and any automated decisions.
  • Accessible formatting, so disclosures are readable by screen readers and on small screens.

Thoughtful UI/UX design keeps these elements noticeable without distracting from the main content.

Everyday uses that need less ceremony

Many common uses of AI carry little risk of misleading anyone, and a site-wide note on your AI use page is usually enough for them:

  • Grammar, spelling and readability suggestions on text your team wrote.
  • Help brainstorming headlines or outlines that a person then writes up.
  • First drafts of routine descriptions, such as product dimensions, that staff check against the real item.
  • Machine translation, provided a fluent person reviews important pages, and with a note where accuracy matters, such as safety or contract terms.

The common thread is human review. If a person who understands the subject checks the result and is willing to stand behind it, the AI’s role is closer to that of any other writing tool. If nobody checks it, disclosure alone will not make it trustworthy.

Avoid both extremes

Some businesses hide AI use entirely, worried that any mention will put customers off. Others label everything, including spelling checks and translation tools, until the labels become meaningless. Both approaches undermine trust over time, in quite different ways.

A better middle path is to disclose where it would change how a reasonable visitor understands the content, and to make sure everything published, however it was made, is accurate and reviewed by someone accountable. Google’s broader point applies here: quality matters more than production method, but honesty about method matters when people would reasonably want to know.

Keep a simple internal record

It helps to know, internally, where AI is used on your site. A short list covering which tools are used, for what, who reviews the output and where disclosures appear makes it easier to answer customer questions, update practices as rules change and keep disclosures consistent when staff change.

Review that list periodically, perhaps alongside other website maintenance, and whenever you add a new tool. Note the date of each review, and record any change to a vendor’s terms, such as how it stores or reuses the content you send it, since that can affect what you should tell customers.

The bottom line

Disclosing AI use on your website comes down to a few practical norms. Disclose where a reasonable visitor would expect to know, especially for realistic images, chat and anything that could mislead. Never use AI to fake reviews or testimonials. Credit accountable people, describe AI’s role plainly and place disclosures where they are relevant. And keep the focus on accurate, genuinely useful content.

If you are planning how AI fits into your site, Motta Industries’ web development team can help you design features and disclosures that keep customer trust at the center.