Why returnable bottles are a counterfeiter’s favorite target

Returnable glass bottles exist because they’re reusable — and that’s exactly what makes them exploitable. A genuine bottle gets collected, cleaned, and refilled at the bottler. But nothing stops a bad actor from doing the same thing outside the official supply chain: collect empty bottles, refill them with a cheaper or unsafe liquid, reseal, and put them back on shelves.

A few things make this scam easy to pull off:
The bottle already looks used. Since returnable bottles are always reused, a worn cap, faded label, or scuffed glass doesn’t raise suspicion the way it would on a “new” product. Consumers expect some wear.
Consumers don’t interrogate old bottles. Nobody asks “why does this look old?” because that’s normal for a returnable — the wrong assumption to make is that old means safe.

Refilling doesn’t require much equipment. A funnel, a cap, and a matching (or salvaged) label are often enough to pass a casual glance.
But counterfeiters aren’t perfect. Fill lines sit slightly off, liquid color is a shade wrong, labels are misaligned or printed on the wrong stock, seals don’t sit quite right. These are small, physical tells — exactly the kind of pattern-level anomaly that machine learning is good at catching, even when a busy shopper or store clerk would miss it.

That’s the gap giwu-ai is built to close: using computer vision to check labels, seals, and fill levels against what a genuine product should look like, so shoppers get a signal before they buy — not after.