Communications of the ACM
Faces Are the Next Target for Fraudsters
In the past year, thousands of people in the U.S. have tried to trick facial identification verification to fraudulently claim unemployment benefits from state workforce agencies, according to identity verification firm ID.me Inc.
Credit: Jamie Chung/The Wall Street Journal
Facial recognition systems increasingly are a target for fraudsters.
Identity verification company ID.me Inc. found more than 80,000 attempts to trick facial identification verification to claim fraudulent unemployment benefits between June 2020 and January 2021.
ID.me's Blake Hall said these attempts involved people wearing masks, using deepfakes, or holding up images or videos of other people.
Veridium LLC's John Spencer said fraudsters sometimes try to carry out "presentation attacks" by using a photo of someone's face, cutting out the eyes and using it as a mask.
Adversa.ai's Alex Polyakov said the algorithms underpinning these systems need to be updated, or the models need to be trained with a large number of adversarial examples, to protect against such spoofing.
From The Wall Street Journal
View Full Article - May Require Paid Subscription
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found