Researchers from the US Develop an AI-Based Tool which can fake Fingerprints and Fool Biometric Systems

A team of scientists from New York University have attained new levels of technological breakthrough. The scientists have created an AI-based tool which has the capability to generate false human fingerprints and probably beat the mechanism of a biometric authentication system, which ultimately will unlock any e-device without any authentication.

Biometric authentication system like finger authentication systems are widely trusted across the world and mostly used in e-devices like smartphones, tabs and several others.

But, the researchers have unveiled a surprising level of vulnerability in these biometric systems.

The researchers created fake fingerprints with the help of a neural network. Out of five people, the fingerprint could fool the biometric authentication process in one person.

Like a master key, which can open any door, similarly, these AI-based “DeepMasterPrints can unlock the majority of the e-devices without any fail.  With the help of AI, these DeepMasterPrints can match prints stored in fingerprint databases.

Nasir Memon, a professor at New York University, earlier led a research process. The researchers used his ideas to further carryout the development. Mr. Memon used the term “MasterPrint” in order to describe the mechanism of the fingerprint-based systems. These systems use partial fingerprints to confirm an identity.

Full prints will be more unique than Partial fingerprints.

The research work of Mr. Memom revealed that MasterPrints are capable of matching many stored partials in a database because of similarities that exist between partial prints. 

In the new development work, the researchers trained a machine learning algorithm to produce synthetic fingerprints as MasterPrints.