Short answer: In a survey commissioned by Auros and run by Talker Research among 4,000 adults (26 August to 1 September 2026), 42% of Australians said they had felt misled by a product advertised using AI, and only 34% felt confident spotting AI product images. When shown a real product whose photo had been enhanced with AI, only 8% of Australians identified it correctly and 74% thought it was entirely AI-generated. For sellers, that means a genuine photo can be doubted as readily as a fake one. The practical answer is a dated record of the camera original, the edited listing image and any AI step, plus a plain note on the listing. A TimeBind dates the files you hold. It does not prove a photo is unedited or that a product matches it.
Last checked 13 November 2026. General information for sellers, marketers and content creators, not legal or advertising-compliance advice. Advertising and consumer rules differ by country, so check the rules where you sell. The survey was commissioned by a company that sells consumer research.
Table of Contents
What the survey found
According to Prompt the Market (28 October 2026), which summarised Auros’s release and press coverage by B&T and Campaign Brief, Auros commissioned Talker Research to survey 4,000 adults, including 1,000 in Australia, between 26 August and 1 September 2026. In Australia, 42% said they had felt misled by a product advertised using AI, 14% said it had happened more than once, 59% said they did not trust AI-advertised products to arrive as described, and 45% said they go out of their way to buy from brands that do not use AI in their ads.
Only 34% of Australians felt confident spotting AI product images, compared with 47% in the UK and 44% in the US. When shown a real 3D waterfall rug whose photo had been enhanced with AI, 8% of Australians identified it correctly and 74% thought it was entirely AI-generated. Auros is the company behind UserTesting, which sells consumer research, so the commissioning interest is worth knowing. We have not seen the full survey.
A related story on X
X’s automated news summary of 11 November 2026 described a thread in which users shared side-by-side video ads for magnetic lashes that, they said, showed AI-generated people with fake urgency banners, and called for clearer disclosure. That is a summary of user posts and can contain errors, and we have not verified the ads, so we do not name the seller. It shows the same worry from the buyer’s side: people increasingly ask whether what they see in an ad is real.
Why this matters to honest sellers
If viewers cannot tell, an honest seller can be doubted and a dishonest one can pass. The only durable answer is to be able to show what you did. That means being able to produce the original photo from your camera, the version you edited and what the edit was, and which, if any, AI tool touched it. It also means telling buyers when an image is AI-assisted, because many platforms and many buyers now ask.
This is not a claim that using AI on product photos is wrong. It is a claim that, given the survey’s numbers, you should be ready to explain what your pictures are.
Why people are searching for this
Measured on 13 November 2026. The Auros survey was covered by Prompt the Market, B&T, Campaign Brief and others in late October, and X’s summary of the AI-people ads thread was updated on 11 November. Google autocomplete in the United States returned “ai generated product images,” “amazon ai generated product images” and “ai ads misleading” on 13 November. Counts keep moving.
Checklist for sellers and creators
- Keep the camera original. Save the untouched file, with its metadata, in a dated folder before any edit. Try the file date metadata inspector to see what a file carries.
- Keep every edited version separately. The listing image should be a different file from the original, so you can show what changed.
- Log edits, including AI steps. One line each: tool, date, what it did. “AI upscale” or “background clean-up” is enough.
- Photograph the actual product you ship. Keep a dated photo of the real item or batch next to the listing image, so the match can be checked.
- Say it plainly on the listing. If an image is AI-assisted or illustrative, say so. See the marketplace listing photo checklist.
- Check Content Credentials if your tools add them. The content credentials checker shows whether a file carries them. Their absence does not mean an image is fake.
- Know the rules where you sell. Advertising and consumer-protection rules differ by country and marketplace. Check them rather than assuming.
What a TimeBind does and does not show
A TimeBind is a dated, checkable record that an exact file existed no later than a recorded time and still matches. The SHA-256 fingerprint is made in your browser and the file is never uploaded. The record is anchored to Bitcoin (a Merkle root in an OP_RETURN). Your first Pooled TimeBind is free; after that, TimeBinds start from US$5, and an ASAP TimeBind is US$15. See how it works.
For a product listing, it can date the camera original, the edited file and an edit log as you hold them. It does not show that a photo is unedited, does not show that a product matches its photo, and does not show whether AI was used. Use it as one dated record beside the originals themselves.
Limits
This page relies on Prompt the Market’s 28 October 2026 article summarising Auros and Talker Research’s survey (with B&T and Campaign Brief cited there), X’s automated news summary of 11 November 2026 and Google autocomplete, checked 13 November 2026. We have not seen the full survey, the ads in the X thread, or any seller’s records. This is not legal or advertising-compliance advice.
Tools that can help
Use the bulk SHA-256 manifest to fingerprint originals and edits together, and try the real or AI product photos quiz to see how hard the call is. For related reading see Etsy print-on-demand design copies.
Frequently Asked Questions
-
How many Australians said they had been misled by an AI-advertised product?
In the Auros and Talker Research survey, 42% said they had felt misled by a product advertised using AI, and 14% said more than once.
-
Can people tell a real product photo from an AI one?
In that survey only 34% of Australians felt confident, and when shown a real product photo enhanced with AI, only 8% identified it correctly while 74% thought it was entirely AI-generated.
-
Who ran the survey?
Talker Research, commissioned by Auros, the company behind UserTesting. It surveyed 4,000 adults including 1,000 Australians between 26 August and 1 September 2026.
-
Can a TimeBind prove my product photo is unedited?
No. It shows exact files existed no later than a recorded time and still match. It does not show a file is unedited or that a product matches it.
-
What should I keep for each product photo?
The camera original, each edited version, a short edit log noting any AI step, and a dated photo of the real product.
Sources checked
- Prompt the Market, “Australians say AI ads have duped them, and most don’t trust them,” 28 October 2026, https://www.promptthemarket.com/australians-misled-by-ai-ads-auros-survey/ (read 13 November 2026)
- Auros, “Auros Study Finds Consumers Wary of AI Retail Ads, but Open to AI as a Shopping Tool,” https://auros.ai/newsroom/ai-shopping-survey-2026 (seen in search results 13 November 2026)
- B&T and Campaign Brief coverage of the survey (as cited by Prompt the Market; not read directly)
- X automated news summary on AI-generated people in ads, 11 November 2026 (Grok-generated, can contain errors) and Google autocomplete (US), checked 13 November 2026
Date your originals, edits and edit log together
Your first Pooled TimeBind is free, with no credit card required. After that, TimeBinds start from US$5, and an ASAP TimeBind is US$15. Click here to start, or see plans and pricing.
