DTA seeks ‘liveness detection’ software to prevent digital ID fraud
- 21 March, 2018 17:16
The Digital Transformation Agency (DTA) has begun the hunt for a key software component for the government’s digital identity program.
The DTA has approached the market seeking “liveness detection” software to help it prevent identity fraud on the Govpass platform.
The intention of Govpass is to allow an individual to verify their identity once and then reuse that digital identity across different government services.
The DTA is building a federated solution that will eventually allow an individual to potentially verify their identity at a number of providers — possibly including trusted private sector entities.
Govpass incorporates multiple components. Last month the DTA released the first component of the Trusted Digital Identity Framework for organisations seeking to provide identity services. The TDIF provides the standards that underpin Govpass.
A central component of Govpass will be an exchange that mediates between members of the identity federation.
Creating a Govpass with the federal government will involve submitting details of a number of personal documents such as a Medicare card or birth certificate.
Photo ID will also be required for most services according to the DTA and an image captured during the Govpass creation process will be checked against the government’s Facial Verification Service to confirm an individual’s identity. The liveness solution will offer an additional level of fraud prevention.
Digital facial verification will allow services that require a high level of identity assuredness to be available online, rather than through a government shopfront, according to human services minister Michael Keenan.
The liveness solution will enable a person to capture an image of their face using a smartphone, tablet or PC. It will include QA functions such as telling an individual that the lighting or background are not suitable as well as crop the image and provide automatic enhancement.
The software will also force a series of checks to be undertaken while an image is being captured. For example it might tell an individual to turn their head in a certain, randomly selected direction or speak a series of words, with multiple images or video captured for analysis.
The software is intended to weed out the use of printed or digital images, video replays or artificial avatars.