Proof of authenticity that AI agents and robots can read.
Photonictag gives every product one tamper-proof identity. Shoppers scan it, returns desks check it, and AI agents verify it on their own, all against a record only the brand can sign.
Photonic Demo
Demo tote bag
- Serial
- DEMO0002
- Made in
- Italy
- Color
- Cognac
- Material
- Calf leather
- Signed by
- Photonic Demo · key b61a…9a58
A real result from our live demo product. Open it.
One identity, readable by people and machines.
No new scanners, no chips. It works with the phone cameras and AI assistants people already use.
The brand signs each item
Every unit gets a record (serial, origin, materials) signed with a key only the brand holds. Nobody can forge or quietly edit it.
It travels with the product
A standard GS1 QR code goes on the item, and the same ID is hidden invisibly in product photos, so images carry proof too.
Anyone verifies in a second
Shoppers scan, warehouses check returns, and AI agents call our verification tools. Each gets a clear verdict with the evidence.
Built for the moments where fakes cost money.
Returns processing
Confirm a return is the exact unit you sold before you refund it. Swapped fakes go to review automatically; genuine items are routed to restock, refurbish or recycle.
Luxury and resale
Marketplaces verify authenticity from a photo and check every seller claim, like origin and material, against the brand's signed record.
AI shopping agents
Agents from Claude, ChatGPT and others connect to our MCP server and verify products on their own. No more trusting whatever a web page says.
Verification as a tool call.
Add Photonictag to any MCP-compatible assistant in one line. Agents get signed facts and a verdict, never a guess.
Developer docsClaude, ChatGPT or any MCP client
MCP server https://tag.photonictag.com/mcp > Is the bag at this link genuine, and is the seller's "made in France" claim true? verify_product_link → verified · Photonic Demo check_listing_claims → origin: contradicts (Italy)
See it on your own products.
We are onboarding a small number of teams in returns processing and luxury resale. A pilot takes about four weeks and needs no new hardware.