Google's current U.S. English Local Services Ads requirements page gives affected advertisers a direct answer. Starting August 2, 2026, Google does not allow AI-edited or AI-generated images in the EU region and New York state.
If a central marketing team supplies Local Services Ads photos to branches in either market, stop using those assets now. Do not replace every photo with a blank profile, though. Google separately says photos can help customers understand a business and may contribute to profile quality.
The operating answer is a real-photo supply chain: capture real service work, preserve the original file, review the asset against the correct market rules, and publish it only to the right location.
Important
Treat Local Services Ads as its own media channel. The current rule names the EU region and New York state, applies to AI-generated and AI-edited images, and does not automatically define what is allowed on Google Business Profile or an owned website.
The broader photo evidence guide explains how real images support local pages and profiles. For brands that need one operating standard across branches, the home services visibility system connects location proof, reviews, listings, and AI search measurement.
What changed on August 2, 2026
Google's Local Services Ads requirements now state that AI-edited or AI-generated images are not allowed in the EU region and New York state starting August 2, 2026. The same page says the photo guidelines do not apply to partner affiliate providers.
Those two sentences set a narrow policy boundary. They do not say that every advertiser worldwide is under the same regional restriction. They do not say that every image on Google Business Profile or a website is governed by the Local Services Ads rule. They also do not explain why Google selected those markets.
Do not widen the claim. Save the live policy URL, the language version reviewed, the review date, the affected market, and the account owner. A New York HVAC branch and a neighboring New Jersey branch may sit inside one advertising organization, but their asset eligibility should not be inferred from the same row in a media library.
Google's page also retains its broader photo restrictions. Photos should be relevant to the work, clear enough for customers to understand, and free from prohibited information such as phone numbers, email addresses, and street addresses. Misleading, duplicate, low-quality, or unrelated images can still fail even when no AI tool touched them.
The rule applies to the finished asset
The policy names two asset states: AI-generated and AI-edited. That means the risk is not limited to a picture made from a text prompt. An authentic job photo can become an AI-edited image after background replacement, generative fill, synthetic object removal, or another AI transformation.
Google's requirements page does not publish a detailed threshold for every editing feature. It does not tell an operator whether each phone enhancement, export setting, crop, denoise tool, or compression step counts as AI editing. A marketing team should not invent certainty where the policy is silent.
For affected locations, keep the untouched original from the camera or phone. Create a record of the branch, service, capture date, photographer, consent status, and destination. Avoid generative transformations. If a necessary production step is unclear, ask Local Services Ads support about that exact step before uploading the image.
This is a provenance problem as much as a creative problem. When the final JPEG arrives in a shared drive without its original, market, and owner, the reviewer cannot prove what happened between capture and upload.
Keep Local Services Ads, Business Profile, and website media separate
Google requires advertisers to add photos directly to the Local Services Ads profile. Its photo guidance says photos from a linked Google Business Profile are not automatically added to Local Services Ads.
Google's separate July 2026 AI labeling update for other advertising products names Google Ads, Display & Video 360, Campaign Manager 360, Merchant Center, and Ads Editor. It does not name Local Services Ads. That is another reason to review each surface under its own current requirements instead of treating one Google policy as a universal rule.
That separation is useful. One image inventory can feed several channels, but every channel needs its own eligibility decision and publication record.
For Local Services Ads, track the ad account, branch, region, review result, and current Google rule. For Google Business Profile, track the profile, location, image category, and its separate media requirements. For the website, track the page, alt text, caption, consent, performance, and visible claim the image supports.
A source photo may be usable on one surface and not another. The correct response is not to copy an LSA restriction into every channel without evidence. It is to maintain a channel-by-channel asset matrix and review the live rule before each release.
Build a real-photo chain of custody
The safest workflow has four gates.
Capture one real job
Ask the branch to photograph real work it is authorized to show. An HVAC team can document a clean equipment installation. A roofer can capture finished flashing or a completed repair detail. A pest control technician can photograph equipment and a non-identifying treatment area. The image should represent the service and location that the ad profile offers.
Preserve the original
Keep the original file before it enters a design tool. Record the location code, market, service, capture date, photographer, and whether customer or property consent is required. Store the original separately from derivatives so a reviewer can trace the final asset back to the camera.
Review policy, privacy, and truth
Check the current Local Services Ads requirements for the account's region. Then check the frame for a phone number, email, street address, license plate, customer face, document, screen, or other information that should not be public. Confirm that a before-and-after pair shows the same job and that the image does not imply a service the branch cannot perform.
Publish and read back
Upload the approved file to the correct Local Services Ads profile, wait for review, and inspect the live profile. Record the asset, branch, upload date, approval state, and next policy-review date. A file sitting in the right folder is not evidence that the right customer-facing profile received it.
Use photos that make the service easier to trust
Google's Local Services Ads photo guidance recommends photos of the team, completed work, before-and-after results, and equipment. Its help page also reports that, in a North American home-services test from December 2023 through January 2024, providers with at least one photo averaged 16% more leads than comparable providers without a photo.
That was a limited historical test, not a promise that one image will increase leads for every advertiser. Google says high-quality images are one part of profile quality, and its ad ranking guidance lists profile quality alongside bid, responsiveness, proximity, verification, hours, and other factors.
The practical standard is simpler than a ranking theory. Use real images that help a customer understand who will arrive, what the team works on, and what completed work looks like. A 75-location plumbing group should not reuse one generic technician photo across every market. A restoration platform should not publish dramatic damage from a stock library as though a local branch handled the loss. A smart-home installer should not generate a perfect control panel that its field team never installs.
The photo does not need to be cinematic. It needs to be clear, truthful, relevant, and attributable to the right operation.
Stop AI editing from entering through the side door
The person uploading the photo may never open an image generator. AI editing can still enter through an agency handoff, a template tool, an automatic cleanup feature, or a contractor who was told to make the image look better.
Give every contributor the same rule for affected markets: no generated scene, no generative fill, no synthetic background, no AI object replacement, and no AI beauty pass. Require the original file with every submission. Make the reviewer record which production steps were used. A realistic-looking final image is not enough.
Also separate editorial illustrations from advertising assets. An article can use a clearly disclosed illustration to explain a workflow. That does not make the illustration eligible for Local Services Ads. The images on this page are editorial artwork; the recommended ad asset is the real, policy-compliant field photo that the branch captured and preserved.
A 14-day rollout for affected locations
Start with New York branches and EU-region accounts, then use the same controls as a readiness standard elsewhere without pretending the regional rule applies globally.
- Days 1 through 3: Export the active Local Services Ads photo inventory for affected accounts. Mark each asset as original camera photo, AI-generated, AI-edited, unknown provenance, or not applicable. Pause or replace prohibited and unknown assets through the account's approved process.
- Days 4 through 6: Assign one central policy owner and one local capture owner per branch. Publish the required metadata, consent, privacy, and original-file rules.
- Days 7 through 10: Capture a small real-photo set at two pilot locations. Use one normal branch and one harder exception, such as a newly acquired brand or shared regional team.
- Days 11 through 14: Review, upload, wait for Google's decision, and read back the live profiles. Fix the workflow before expanding to the remaining locations.
Do not treat a successful pilot as permanent approval. Save the policy-review date and schedule another check when Google changes the requirements page, the account enters a new market, the media toolchain changes, or a photo is disapproved.
What to monitor after launch
Track policy and performance separately.
The policy record should show which rule was reviewed, where it applied, who approved the asset, and whether Google accepted it. The performance record can show profile views, leads, booked jobs, and cost by location, but it should not imply that a photo alone caused the result.
Keep that channel record separate from AI search visibility. The AI search traffic tracking guide separates visibility, citations, visits, and booked demand. A location-level visibility check can show which businesses answer engines recommend, but it does not validate Local Services Ads eligibility or photo approval.
Watch for disapprovals, removed images, assets with unknown provenance, branches without usable real photos, and agencies that cannot provide the original. Recheck the customer-facing profile after every upload. If Google changes the regional scope or defines AI editing more precisely, update the asset matrix before the next batch.
For multi-location service brands, the win is not a clever image workaround. It is a repeatable system that proves which branch performed which work, keeps the original intact, and publishes the right evidence under the right market rule.
Sources
- Google Local Services Ads: Business screening and verification requirements. Current AI-generated and AI-edited image restriction for the EU region and New York state, partner affiliate exception, and broader photo restrictions.
- Google Local Services Ads: About Local Services Ads with photos. Photo purposes, direct LSA upload requirement, review process, recommended subjects, and the December 2023 through January 2024 North American home-services test.
- Google Local Services Ads: About ad rankings. Profile quality, images, responsiveness, verification, proximity, bid, hours, and other documented ranking inputs.
- Google Local Services Ads: Edit your ad or business information. Current profile photo management path and examples of real employee, work, and before-and-after images.
- Google Ads: Updates to AI labelling requirements. July 2026 AI-label support for named advertising products outside Local Services Ads.
Amadeus Peterson is the CTO & Co-Founder of Cheers, the local search platform for multi-location service businesses.