AI Scales Luxury Authentication – But Where Are the Security Features?

Artificial intelligence is no stranger to luxury authentication. For several years, resale platforms and brand protection providers have been using image recognition, database matching, and machine learning to support the identification of counterfeit goods. What is changing now is not the presence of AI, but its role.

At The RealReal, that role is becoming operational. Inside its New Jersey warehouse, in the US, which handles millions of second-hand luxury goods each year, a system known as Athena is no longer simply assisting authenticators – it is actively shaping workflow decisions.

Rather than sitting alongside human judgement, Athena now determines how items move through the system. It analyses incoming goods, assigns a counterfeiting risk score, and decides whether an item can bypass manual inspection or requires expert review. In doing so, it transforms authentication from a uniformly applied process into a prioritised, risk-based model.

And yet, for all the sophistication of this system, there is a striking omission at its core.