Researchers continue to improve on a technique for embedded crafted outputs into machine-learning models, an anti-copying technique originally thought up by adversarial researchers.

from Dark Reading https://ift.tt/aLWo583
via IFTTT