Somewhere in your portfolio is a customer you love.
Pays early. Never calls support. Never disputes a charge. Asks for a higher limit only after years of spotless tenure, and your model grants it gladly, because every signal says: more of this, please.
Then one day they max out every line of credit, all at once, and disappear. There is no person to call. There never was.
That is a synthetic identity, and it beats you the way a great method actor beats an audience.
It does not play the part of your best customer. It becomes your best customer.
There is no moment where the mask slips, because there is no face under the mask, no real self waiting offstage to break character. The performance is the whole person.
Our fraud systems are built for bad actors who are also bad actors. People acting badly, rushing, fumbling, tripping a velocity rule. We watch and wait for the tell. A method actor has no tell. The clean payments, the low support burden, the patient utilization, none of that is cover for the fraud. It is the fraud, performed in full.
There is no face under the mask, no real self waiting offstage to break character.
This is why synthetic identity is a product problem before it is a model problem. You cannot wait for the character to break, because it never will. You verify there is a real human behind the role before the curtain goes up, and you keep asking through the run of the show, instead of letting a flawless performance earn permanent trust.
The sloppy fraudster forgets his lines. Your models were built for him.
The one you should fear knows the part better than your real customers do.
Shyam Menon is a product leader specializing in fraud and identity in financial services. This is one of a series of framework posts on how to think about fraud prevention, identity, and AI products in regulated industries. He writes at shyammenon.com.