Buying a watch secondhand is exciting. It can also be nerve-wracking. The superclone market has become disturbingly good. Franken watches, assembled from mixed genuine and counterfeit parts, are even harder to catch. And for most buyers, getting a professional authentication means sending the watch away for days, sometimes weeks, and paying hundreds of dollars before you even know what you're dealing with.
We wanted to give people a first line of defense. Something fast, something free, something that lives right where your watch data already is.
What the screener actually does
The authenticity screening runs a multi-layered check against the data you've already registered for your watch. No photos required. No shipping. No waiting.
Layer 1: Reference validation. We cross-check your serial number format against known patterns for your brand. A Rolex serial from 2022 should be an 8-character randomized alphanumeric string. An Omega from the 78-million range should match a specific era. If your serial format doesn't fit the brand's known patterns, the system flags it immediately.
We also verify the reference number. If you say it's a 126710BLNR, we confirm that reference exists, that it's a GMT-Master II, and that the production year you've entered falls within the window that reference was actually manufactured.
Layer 2: AI analysis. The interesting part. We send your watch data through an AI model trained on brand-specific authentication knowledge. It analyzes serial patterns, reference consistency, production year alignment, known counterfeiting targets, and franken watch indicators. Each area gets its own confidence score.
For example, the AI knows that the Rolex 126710BLNR (the "Batman") was introduced in 2019 and is still in production. It knows that post-2010 Rolex serials are randomized. It knows which references the superclone factories target most aggressively. And it weighs all of this against your specific combination of data points.
Layer 3: Expert review. For cases that need a human eye, verified watchmakers on the platform can review flagged results and add their professional assessment. This layer is rolling out to premium members soon.
What it doesn't do
This is important: the screener works with data, not photos. It can't inspect your dial printing, check movement finishing, or evaluate lume application. It tells you whether your numbers add up. That's a meaningful first step, but it's not a replacement for hands-on inspection.
The system is clear about this. Every result comes with a confidence score, not a binary yes/no. And each finding includes a recommended next step, so you know exactly what to do if something needs closer attention.
How to use it
If you have a watch registered on Aikakone, go to its detail page. You'll see the Authenticity Screening card in the sidebar. Click "Run Screening" and wait a few seconds.
You'll get back a status (ranging from "Authentic" to "Likely Counterfeit"), an overall confidence percentage, and a breakdown of individual findings. Each finding can be expanded for details and a specific recommendation.
Don't have a watch registered yet? Start at aikakone.io/authenticate to learn more.
Why confidence scores instead of pass/fail?
Because authentication isn't binary. A watch can have a valid serial format and a confirmed reference number but still be a superclone. Data screening catches obvious red flags and validates what can be validated. The confidence score tells you how much of the picture we can actually see.
An 82% "Likely Authentic" on a Rolex GMT-Master II means the data checks out across multiple dimensions, but conclusive authentication still requires physical inspection. That's not a hedge. That's accuracy.
What's next
We're working on photo-based analysis as a future layer. Upload images of your dial, caseback, and movement, and the AI compares visual details against known authentic references. That's a bigger technical challenge, but it would let the screener catch things that data alone can't.
For now, the data-based screening is live and free for every registered watch. Try it on your collection and see what comes back.
