Compliance and Privacy11 min read

Can AI Write Google Reviews for Your Business?

Writing fake Google reviews with AI for your business is banned by the FTC rule. See what you may automate, what you may not, and the $53,088 penalty.

Can you use AI to write Google reviews for your business? If you run a Las Vegas restaurant, dealership, law firm, or real estate office, the answer is no for the reviews themselves and yes for nearly everything around them. Since October 21, 2024, the FTC's Consumer Review Rule has banned businesses from writing or buying reviews that misrepresent who the reviewer is or what they experienced, which the FTC said in August 2024 covers AI-generated fake reviews. Knowing violations can draw civil penalties up to $53,088 each, the figure set in the FTC's February 2025 civil penalty adjustment.

What changed is enforcement. The FTC sent warning letters to 10 companies in December 2025, and in May 2026 the Justice Department and the Illinois Attorney General sued a Chicago company over thousands of fake business listings and fabricated five-star reviews. This is not legal advice. It is a map of the rule's text and Google's policy, so the questions you bring to your lawyer are sharper.

Can you use AI to write Google reviews? Two rulebooks give the same answer

The federal rule is 16 CFR Part 465. Section 465.2 makes it unfair or deceptive for a business to write, create, or sell a review that materially misrepresents that the reviewer exists, that they used the product or service, or what their experience was. It also bars buying or posting reviews the business knew or should have known were false.

The rule never mentions software. AI is a tool, and the test is the output. A model that produces a glowing paragraph signed by a customer who does not exist has created a review from a nonexistent person, the exact case the FTC's August 2024 final rule announcement named as AI-generated fake reviews. A model that helps you word a reply to a real customer has done nothing the rule addresses.

Google's Maps user contributed content policy, which we read on September 25, 2026 (the page shows no date), lands in the same place. It bans content not based on a real experience, reviews paid for directly or in kind, and content posted from multiple accounts by or at the request of one person. It names no rule on AI-generated text. So the question is never whether AI touched the words. It is whether a real customer had this experience and chose to say it.

The rule also marks the safe path. Section 465.2(d) excludes reviews that came from a business making generalized solicitations to purchasers to post reviews about their experiences, and reviews that appear because the business merely hosts them. That carve-out is the foundation of every legal workflow below.

Ten review tasks, sorted by what the FTC rule and Google allow

TaskFTC rule (16 CFR 465)Google Maps policySource
Generalized review request to every customer after servicePermitted, 465.2(d) carve-outAllowed; no pressure on premises, no payment16 CFR 465.2, Google Maps content policy
Requests sent only to happy customersUnprotected: outside the 465.2(d) carve-outNot named; staff may not solicit specific content16 CFR 465.2, Google Maps content policy
Discount for any review, any ratingPermitted, 465.4 reaches only sentiment-tied incentivesBanned16 CFR 465.4, FTC rule Q&A
Discount for a five-star reviewBanned, 465.4Banned16 CFR 465.4
Drafting the review text for a customerHigh risk, 465.2Must reflect a real experience16 CFR 465.2
AI-written reviews from fake personasBanned, 465.2BannedFTC final rule release
Employees or relatives posting reviewsBanned when it misrepresents reviewer or experience, 465.2Banned if not a real experience16 CFR 465.2
Drafting owner replies with AIPermitted, not addressed by Part 465No AI rule namedGoogle Maps content policy
Threatening a reviewer to get a removalBanned, 465.7Paying for removal banned16 CFR 465.7
Hiding low ratings on your own testimonials pageBanned, 465.7Not applicable, your own site16 CFR 465.7

Two rows sound harmless and are not. Sending requests only to happy customers, often called review gating, is common in reputation software: a text asks how the visit went, and only the top ratings get the Google link. The 465.2(d) carve-out protects generalized solicitations, and a filtered request is not generalized. Neither source names gating as a standalone violation, so we call it unprotected rather than banned. That is still a weak place to build.

Drafting the review for a customer is the other trap. Owners frame it as a favor: the customer is busy, so the business writes three sentences and the customer pastes them. But 465.2 targets a business that writes or creates a review misrepresenting the reviewer's experience, and the words describing that experience are now yours. Send the link, not the text.

The last row matters for your website. Section 465.7 bars a business from implying its testimonials page shows most or all reviews while suppressing low ratings. Withholding is allowed only under criteria applied equally regardless of sentiment, such as defamatory, clearly false, or likely fake content.

Section 465.4 bans incentives "in exchange for, or conditioned expressly or by implication on" reviews expressing a particular sentiment, positive or negative. The FTC's November 2024 questions and answers page states the converse: incentives are permitted when there is no express or implied requirement that the review take a side.

That makes a neutral discount legal under the federal rule. It does not make it allowed on Google. The Maps policy prohibits merchants from offering payment, discounts, or free goods or services in exchange for posting reviews, at any rating. For an owner whose reviews live on Google, the platform policy is the tighter rule.

Take a Henderson taqueria that prints an offer of 10% off the next visit for any Google review on every check. Under 465.4 that is permitted, because no sentiment is required. Under Google's policy it is a review paid for in kind. Change the wording to promise free dessert for five stars and it fails both rulebooks. The phrase "by implication" also catches the softer version: a reward offered only to guests who say they loved the meal.

Why owners are tempted shows up in BrightLocal's Local Consumer Review Survey 2026, published February 2026 from 1,002 US adults surveyed through SurveyMonkey. It found 31% of consumers will only use a business rated 4.5 stars or higher, up from 17% in 2025, and 68% require four or more stars. A restaurant at 4.3 falls outside the filter for nearly a third of searchers.

The stakes run the other way. As a ceiling illustration only, assume a regulator counted each of 30 sentiment-bought reviews as one violation: 30 times $53,088 is $1,592,640. Courts set actual penalties, and the FTC's Q&A ties them to knowing violations, so the real figure could be far lower. The ratio is the point: a 10% discount buys a review, and the maximum exposure per review is tens of thousands of dollars.

Enforcement moved from rule to warning letters to a federal lawsuit

ItemDateFigureSource
Maximum civil penalty per violationEffective Jan 17, 2025$53,088, up from $51,744FTC 2025 penalty adjustment
FTC warning lettersDec 22, 202510 companiesFTC warning letters release
Premium Home Service complaintMay 11, 2026Thousands of fake listings since at least 2018FTC and Illinois complaint release
Consumers wanting consequences for fake reviewsFeb 202697%; 37% want financial penaltiesBrightLocal fake reviews breakout
Research detection accuracyMay 202693% Amazon, 91% YelpTech Xplore detection report

The December 22, 2025 letters cited fake reviews, paid-for sentiment, undisclosed insider reviews, manipulated social metrics, and review suppression, which covers most of the banned rows in the first table. Letters are not penalties, but they show what FTC staff are looking for.

The May 11, 2026 complaint shows what a full case looks like. The Department of Justice, acting for the FTC, and the Illinois Attorney General sued Chicago-based Premium Home Service and its owner, alleging it built thousands of profiles for home-repair companies that did not exist, diverted tens of thousands of consumers to them, and posted fabricated five-star reviews to dilute legitimate one-star ones. The charges include writing, buying, or selling fake reviews and using reviews from employees and relatives.

Detection is improving. Tech Xplore reported in May 2026 that a University of East London model, published in FinTech and Sustainable Innovation, combined transformer language analysis with behavioral clues, such as whether a review's emotional tone matches its star rating, and reached 93% accuracy on Amazon data and 91% on Yelp reviews. Dataset sizes were not stated, and it is a research result, not a platform's live filter. It does show that fluent AI text can still be flagged by the mismatch around it.

Consumers want enforcement. BrightLocal's February 2026 breakout of the same survey found 97% think businesses posting fake reviews should face consequences, 57% support removal from review platforms or Google search results, and 16% favor criminal charges. BrightLocal also notes AI is producing authentic-sounding reviews at scale.

What AI can run in the review workflow without crossing a line

The legal workflow has three automated pieces. When your point of sale, dealer management system, case tool, or CRM marks a job closed, every purchaser gets the same request with the Google link and no rating screen in front of it. Anything a customer sends you directly, including complaints, gets tagged and routed to the right person. And AI drafts a reply to each new public review, which a human approves before it posts.

Replies matter because the same BrightLocal survey found 19% of consumers expect same-day owner responses. Nothing in Part 465 governs how you word your own reply, and Google names no AI rule. The guardrail is accuracy: a draft must never invent a refund, a promise, or a fact. This is the kind of build we cover under back-office automation, and a reply drafter trained on your policies is a small custom AI assistant project.

Here is the volume math for a different size of business. Take a two-attorney Las Vegas firm with 12 Google reviews. BrightLocal found 47% of consumers will not use a business with fewer than 20 reviews, so the firm needs 8 more. Assume it closes 30 matters a month and 15% of clients who get a request post a review. That is 30 times 0.15, or 4.5 reviews a month, so 8 reviews takes about 1.8 months.

The flow has to keep running, because 74% of consumers look for reviews written in the last three months. If the assumed 15% response rate is really 5%, the same 30 matters yield 1.5 reviews a month, and 8 reviews takes over five months. The fix is asking every client, consistently, never a filtered list or a review written for them.

Where this map does not hold: other platforms, state rules, and survey limits

This is not legal advice. Part 465 is a federal rule, and Nevada law, State Bar advertising rules, and dealer licensing rules sit on top of it; none are covered here. Yelp and other platforms publish their own policies, which we did not review, so a practice allowed on one may be banned on another.

The Google page carries no date and can change; our reading is from September 25, 2026. The $53,088 figure is a statutory maximum, not a typical outcome, and we found no 2026 adjustment as of September 2026. The FTC's Q&A ties penalties to knowing violations, so what you knew matters.

The consumer numbers are stated preferences from one online survey of 1,002 adults, not observed behavior, and Las Vegas buyers may differ. The detection result comes from one study with undisclosed dataset sizes. The gating row is our reading of the carve-out; if a vendor sells gating as compliant, ask which section they rely on.

Questions owners ask

Can I use ChatGPT to reply to my Google reviews?

Yes. Nothing in the FTC's Consumer Review Rule governs how a business writes its own replies, and Google's contribution policy as read in September 2026 names no rule on AI-generated text. The risk is accuracy, not authorship. Have a person approve each draft so it never promises a refund or states an unchecked fact, and never reply with a threat or public false accusation, which 465.7 bans.

Is it illegal to offer a discount for a Google review?

Under the FTC rule, a discount for any review is permitted as long as there is no express or implied requirement of a positive or negative sentiment. A discount tied to five stars is banned under 465.4. Google's policy is stricter: it prohibits payment, discounts, or free goods for posting reviews at all. For Google reviews, the practical answer is not to offer one.

Can my employees leave Google reviews for my business?

Not as if they were customers. Section 465.2 bars procuring reviews from employees, managers, agents, or their relatives that misrepresent who the reviewer is or their experience, and the Premium Home Service complaint charged exactly that. Google bans content not based on a real experience and bars staff from soliciting a set number of reviews. Have staff send the request link to customers instead.

What to do this week

  1. Test your review request. Send yourself one from your current tool. If anything screens your rating before showing the Google link, or the message asks for five stars, replace it with one generalized message sent to every purchaser.
  2. Pull Google review incentives. Remove any discount, freebie, or contest tied to posting a Google review from receipts, signs, and email. Everywhere else, drop any reward linked to a positive rating.
  3. Fix your testimonials page and reply process. If low ratings are hidden, publish them or adopt a sentiment-neutral rule, such as removing only defamatory or likely fake reviews, applied to every rating. Then set AI reply drafts to require human approval before posting.

If you want a second set of eyes on your review workflow, the free 15-minute audit on our back-office automation page covers it.

Drafted with AI assistance, researched, edited, and fact-checked by Elias Musleh on November 1, 2026.

Free 15-min automation audit

Want this handled for you?

We build the automation, run it, and hand you the numbers. Book a fifteen minute call and we will tell you straight whether it is worth doing for your business.

Or reach us directly: vendors@vegasbusinessai.com 702.826.9055