AI-linked fashion scams drive complaint surge: what customer support needs to do next
Citizens Advice recorded almost 18,000 fashion complaints last year, up 21% year-on-year. The charity says increasing use of AI in product imagery is helping scammers sell items that look nothing like the photos.
Support teams are feeling this. More orders arrive late, more items are poor quality, and more buyers discover they've unknowingly ordered from overseas sellers with costly return hurdles.
The data at a glance
- 18,000 complaints (+21% YoY). Citizens Advice assisted a fashion buyer every seven minutes.
- 82% online orders (14,487). 14% in-store (2,569).
- Women's clothing: 48% of all complaints (8,508). Men's: 20% (3,523).
- Top issues: faulty goods (18%), delivery failures/delays (13%), trouble returning (12%), breach of contract (9%), poor service (6%).
- About 1 in 13 cases involved scams, often with sellers posing as UK-based but actually overseas.
How the scams play out
Common pattern: an ad shows premium-looking items, often with AI-polished images. The buyer thinks it's a UK brand. The product arrives weeks later, looks and feels cheap, and returns are only accepted to an address overseas-at the customer's expense.
One shopper paid £35 for a jacket "from London," received a different material and color, plastic buttons instead of metal, and a demand to return it to China at her own cost. Her bank refunded via dispute; the seller later followed.
Why this matters for support leaders
Scams inflate contact volume, refund pressure, and chargeback risk. They also drain agent time with back-and-forth over "not as described," origin confusion, and return costs.
Your edge: tight triage, clear scripts, and a consistent stance on misrepresentation and overseas returns. The faster you identify a scam pattern, the lower your handling time and losses.
A fast-response playbook you can deploy now
- Front-door triage: Tag tickets by source (social ad, marketplace, website), product category, and suspected "not as described." Route "scam suspected" and "overseas return" to a specialist queue.
- Evidence checklist: Ask for order ID, product photos next to packing slip, screenshots of the ad/product page, and any return instructions received. This shortens investigations.
- Seller origin check: Confirm the shipping origin and returns address immediately. If the ad or site claimed "UK-based" but returns go overseas, treat it as potential misrepresentation.
- Refund paths: If paid by debit/credit card, guide customers to bank disputes, including chargeback or Section 75 where eligible. Keep a one-pager with issuer-specific steps.
- Overseas return costs: If an item is "not as described," push for the seller to cover return shipping or provide a prepaid label. Where that fails, help customers escalate via their bank.
- Delivery delays: For extended delays beyond the stated window, offer options: cancel and refund, or partial credit. Set clear SLAs.
- Ad escalation: Capture the ad URL and report it to the platform. Block repeat offenders and share intel with fraud/compliance.
- Customer education: Provide a short checklist in your auto-reply or help center on spotting AI-altered images and verifying seller location.
Spotting AI-altered product photos (train agents to look for this)
- Overly airbrushed fabric with no natural creases or inconsistent textures across panels.
- Distortions around zips, buttons, stitching, or hands; reflections that don't match materials.
- Models with mismatched proportions, repetitive patterns, or odd lighting.
- Product images that differ from customer photos in color tone and hardware quality.
Ready-to-use micro-scripts
- Not as described (first reply): "Thanks for the photos and the product page screenshot. Based on what you've sent, this differs from what was shown. We can offer a full refund and will cover return shipping. If you paid by card, you can also request a bank dispute (chargeback/Section 75 where applicable). Which option do you prefer?"
- Overseas returns disclosed late: "The listing indicated UK-based operations, but the return address is overseas. That's not what was presented. We'll process a refund without asking you to ship at your expense. You'll receive a label or confirmation within 24 hours."
Policy guardrails to reduce repeat contacts
- Define "not as described" with clear examples and guarantee a no-cost return or instant refund for verified cases.
- Auto-approve refunds for flagged sellers or SKUs with a high misdescription rate.
- Set a firm timeline: if no dispatch within the promised window, allow instant cancellation and refund.
- Publish a one-page "Shop Safe Checklist" in your help center and include it in order confirmation emails.
KPIs to watch weekly
- Percentage of tickets from social ads and unknown brands.
- "Not as described" rate by SKU and seller; repeat offender list.
- Average time to refund for scam-related cases.
- Chargeback submission and win rate; refund cost per case.
- Shipping origin mismatch rate (claimed UK vs. actual).
What experts recommend
Consumer advocates urge shoppers to pause before buying, verify URLs, and be wary of heavy discounts that feel too good to be true. They also advise avoiding direct purchases through social ads and sticking to secure payment methods.
- Citizens Advice guidance on consumer rights and refunds: citizensadvice.org.uk/consumer
- URL and retailer checks: Get Safe Online
Level up your team's AI-readiness
Equip agents to spot AI-synthetic product imagery, triage faster, and reduce refunds lost to scams. For structured upskilling on triage, automation, and fraud-aware support workflows, see the AI Learning Path for User Support Specialists.
Bottom line
AI-polished listings are inflating fashion complaints and disputes. Tighten triage, standardize refunds for misdescribed goods, and coach teams to spot image red flags. Do that, and you cut handling time, protect customers, and keep refund costs under control.
Your membership also unlocks:
