Etsy on October 22, 2026 set out a new standard for artificial intelligence on its marketplace, requiring sellers to include at least one unaltered photograph of the actual item against a real background in every listing, and barring listings generated by bots without human involvement. The company also said functional guides such as knitting patterns and recipes can no longer be created with AI, and that listings for them must show a real result made from the instructions.

The changes were published on Etsy News in a post titled "Creativity and Trust in the Age of AI," written by James Ossman, Etsy's vice president of Customer Operations. They amend Etsy's Listing Image Requirements, part of the House Rules that govern what may be sold on the platform. The post does not give a date on which the new requirements take effect.

In Short

Etsy now says every product listing needs at least one real, unedited photo of the item, and it will not allow shops run entirely by bots or AI-written knitting patterns and recipes. This matters because buyers on Etsy are often paying for something handmade or one-of-a-kind, and AI images can make an item look different from what arrives. Sellers who use AI tools can still do so, but the AI images must match the real item and AI use still has to be disclosed.

What changes for sellers

The post lists three policy changes, each tied to the Listing Image Requirements.

The first concerns photography. According to Etsy, "sellers will need to include at least one unaltered photo of the actual item against a real background. Any supplemental AI-assisted listing images must accurately represent the item a buyer will receive." AI-generated or AI-edited images are therefore not prohibited outright. They are permitted as supplements, provided at least one genuine photograph anchors the listing and the extra images do not misstate what the buyer gets.

The second targets automation. Etsy said it is "making clear that bot-generated listings without human involvement are not allowed, nor is AI use that misleads buyers or creates fraudulent inventory." The phrase "fraudulent inventory" points at a specific failure mode on open marketplaces: listings for items that do not exist in the form shown, produced at scale and often with generated imagery.

The third applies to a single product category. According to the post, "Functional guides, such as patterns and recipes, which require buyers to put in effort to achieve a finished product, cannot be created using AI. Listings for functional guides must also show a real result made using the instructions." This is the only category in the announcement where AI creation itself is ruled out, rather than merely conditioned on accuracy or disclosure. A buyer who purchases a crochet pattern invests hours of work before discovering whether the instructions function; a generated pattern that produces something unlike its preview photo wastes that effort, which is presumably why Etsy singled the category out.

One existing obligation stays in place. "With these changes, sellers are still required to disclose when AI is used to create an item," Ossman wrote. The post links to a Seller Handbook article on the updated policies. That handbook entry is titled "Important updates to AI use on Etsy," but the post itself does not spell out transition periods, penalties for individual violations, or how existing listings will be treated.

Why Etsy rejected a blanket ban

Etsy framed the approach as a deliberate middle path. According to the post, nearly six million "creative entrepreneurs" sell on Etsy, and nearly 90% run their shops on their own, a scale that puts Etsy among the largest concentrations of independent sellers in the creator economy. Tools that handle administrative work, the company argued, can free up time for the creative side of the business.

"The challenge is the vast array of tools available and the different ways sellers may use them," Ossman wrote. "Some uses help sellers bring creative ideas to life or make the tedious tasks involved in running a business more efficient; others can obscure whether there was meaningful human involvement, or misrepresent what an item actually is. A blanket ban on AI tools would treat those uses as if they were the same."

The commitment at the centre of the post is stated in two sentences: "AI should support our sellers' creativity and craftsmanship, not replace them. When shopping on Etsy you should always be able to understand the role sellers play in every item."

The post is unusually candid about the state of the problem. "We know we have work to do to uphold this standard on Etsy," Ossman wrote, and later: "This work is hard, and we don't have every answer today." He also acknowledged pressure from the platform's own community. "We've heard from our community that they want to hear more from us about AI's place on Etsy. That feedback is fair."

What the post does not include is any measurement. There is no figure for the share of listings that use AI imagery, the number of bot-generated listings Etsy has removed, or the volume of buyer complaints about items that did not match their photos. Without those numbers, the scale of the issue the rules are meant to address cannot be assessed from the announcement alone.

Enforcement: detection first, removals later

Policy is only one of three components Etsy described. The others are enforcement and changes to the shopping experience.

On enforcement, the company said it is "leveraging technology to better identify misleading imagery and fraudulent listings." The language that follows is carefully staged: "As those tools continue to get stronger and more accurate, we'll carefully test enforcement and continue to expand removals of violative listings in the coming months." The wording indicates that automated removal is not yet running at full scale. Detection accuracy matters here because false positives would fall on independent sellers, the very group the policy is meant to protect, and Etsy appears to be signalling that it will not switch on aggressive takedowns until the classifiers are reliable.

The post does not say which technologies Etsy will use. It does not mention provenance standards such as C2PA content credentials or invisible watermarks, which some generators embed in their output and which platforms have begun reading to apply labels automatically. Nor does it say whether the requirement for an "unaltered" photo will be checked by software, by human review, or primarily through buyer reports.

Those reports are the third enforcement lever. Etsy said it is "making it easier for users to report listings they believe include misleading AI imagery, which will help us better understand how they perceive AI misuse." The second half of that sentence is notable: the reporting tool is described as a research instrument as much as a moderation one, a way of learning where buyers draw the line.

Changes to how buyers see listings

Rules alone, the post concedes, "can't help buyers find items they'll love or understand the role humans played in creating them." Etsy said it is exploring ways to give buyers more context, citing two examples: "richer reviews" and content "highlighting sellers' creative processes." Neither comes with a launch date or a description of how it would work, and both are presented as areas of exploration rather than commitments.

The post closes with a statement of intent rather than a timetable. "These changes are an important first step, not the finish line," Ossman wrote. "Doing this work well will take time, but we feel a great deal of urgency and are committed to making meaningful progress that our community can feel."

How this fits the wider shift on AI labelling

Etsy's update arrives during a year in which platforms, regulators and ad systems have converged on disclosure and provenance for generated content, though each from a different angle.

In Europe, the regulatory deadline has already passed. The European Commission published guidelines and a final Code of Practice on AI-generated content transparency under Article 50 of the AI Act, with obligations taking legal effect on August 2, 2026. As PPC Land reported, deployers such as publishers and advertisers must visibly label deepfakes, and violations can bring fines of up to EUR 15 million or 3% of worldwide annual turnover. Etsy's post does not mention the AI Act or any other law, and it frames the changes entirely around buyer trust.

Advertising platforms moved earlier in the summer. Google's July 9, 2026 changelog introduced a setting that lets advertisers disclose AI-generated or AI-edited creative across Google Ads, Display & Video 360, Campaign Manager 360, Merchant Center and Ads Editor, with visible on-ad overlays for campaigns targeting the EU, India or New York. Merchant Center's inclusion is the closest parallel to Etsy's change, since both concern product imagery shown to shoppers. The difference is that Google's system centres on labelling, while Etsy now requires a real photograph regardless of any label.

Content platforms have taken a third route, marking content by how it was made. Only days before Etsy's post, YouTube Music began labelling songs made primarily with generative AI on October 14, 2026. Facebook in July 2025 moved against accounts that repeatedly share unoriginal content, after acting against 500,000 accounts in the first half of that year. The underlying concern is the same one Etsy describes: generated material can be produced at a volume that drowns out human work. An analysis by video tool Kapwing of 500 consecutive videos in a fresh account's feed found 21% were AI-generated slop, a measure of how quickly feeds can fill when no friction is applied.

Marketplaces also face scrutiny over what they allow sellers to list. The European Commission in May 2026 fined Temu EUR 200 million under the Digital Services Act for failing to properly assess the risks of illegal products on its platform. That case concerned product safety, not AI, but it illustrates the regulatory exposure that comes with listings that misrepresent what buyers receive.

Why it matters for marketers and sellers

For sellers who advertise through Etsy Ads or drive traffic from Google Shopping, Pinterest and social channels, the listing image is the ad. A requirement that at least one image be an unaltered photo against a real background constrains the cheapest form of product photography now available, where a generator places a product mock-up into a styled lifestyle scene. Sellers who rely entirely on such images will need to shoot at least one real photograph per listing, and print-on-demand and digital-download sellers, whose items may never be physically produced before sale, may find that requirement harder to meet. The post does not address how the rule applies to digital products other than functional guides.

The ban on AI-created patterns and recipes removes a category that had been easy to mass-produce. Digital downloads require no inventory or shipping, so generated guides could be listed in bulk at negligible cost. Under the new rules, each such listing must also show a real result made using the instructions, which adds a physical production step to every item.

For buyers, and for the advertisers whose own trust signals depend on platform-level credibility, the update raises a question the post leaves open: how quickly and how accurately Etsy can detect violations among listings from nearly six million sellers. Etsy has said removals will expand "in the coming months." The pace of that expansion, and the accuracy of the tools behind it, will determine whether the rules change what buyers actually see.

Timeline

Summary

Who: Etsy, through a post by James Ossman, vice president of Customer Operations, affecting the nearly six million sellers on the marketplace and the buyers who shop there.

What: Etsy updated its Listing Image Requirements to require at least one unaltered photo of the actual item against a real background, barred bot-generated listings without human involvement and AI use that misleads buyers, and banned AI-created functional guides such as patterns and recipes. AI disclosure remains mandatory, and Etsy plans stronger detection, easier reporting of misleading AI imagery, and expanded removals.

When: The post was published on October 22, 2026. It gives no effective date for the new requirements, and says removals will expand in the coming months.

Where: On Etsy's global marketplace, with the policy set out in the House Rules and a Seller Handbook article.

Why: Etsy said buyers need to understand the human involvement behind what they buy, and that a blanket AI ban would treat helpful and misleading uses of AI the same way. The move follows community requests for clearer guidance and comes as regulators and platforms have expanded AI labelling rules through 2026.