Ah but the problem is humans, not the AI! Best get rid of those ASAP.

/s

  A Facewatch spokesperson also said the their technology was "not at fault" in this case.

  "A correct alert was sent to the retailer, but was subsequently subject to human error in the way it was handled in store," they said. 

AKA "We correctly notified the store that this guy looked a lot like some other guy" therefore "it's the stores fault that they acted on that."

What value is FaceWatch if it's an error to take it seriously?