Monday, April 15, 2024
HomeSECURITYApple has taught its technique to imitate the human voice with high...

Apple has taught its technique to imitate the human voice with high accuracy

-


Apple has taught its technique to imitate the human voice with high accuracy

Will enterprising scammers be able to abuse the new feature?

Yesterday Apple released A quick look at the new features for iOS 17 coming later this year. Two interesting features called “Live Speech” and “Personal Voice” immediately attract attention.

“Live Speech” is a feature aimed at increasing inclusiveness and facilitating access to full-fledged voice calls for people who have certain problems with the speech apparatus caused by certain diseases. Live Speech allows, for example, during a work call or a personal conversation via FaceTime, to enter text from the keyboard into a special field, and it will be voiced by an artificial speech synthesizer, which will be loudly and clearly transmitted to the rest of the participants in the conversation.

“Personal Voice” is a feature that allows a person to digitize their own voice and use it in conjunction with “Live Speech” to speak to other people in their own voice on the same audio calls without saying a word. The feature is also aimed at people who may lose their own voice over time.

“For users who are at risk of imminent loss of speech due to diagnosed ALS (Amyotrophic Lateral Sclerosis) or other diseases that may gradually affect the ability to speak, Personal Voice is a simple and safe way to create an artificial voice similar to their own.” , — declare in Cupertino

According to Apple, it will be enough to read text prompts aloud on an iPhone or iPad for 15 minutes so that, thanks to machine learning, the system can create a voice that is as close as possible to a real human voice.

Undoubtedly, Apple’s efforts to make the lives of its users more comfortable are commendable, but such a feature, built into the regular iPhone by default, will certainly open up new horizons for fraudsters to interact with relatives and friends of potential victims.

In early April, we reported that the rapid development of neural networks has already led to an increase in fraudulent calls with fake voices. Moreover, such stories are often shocked by the details .

It is not entirely clear whether it will be possible to use the new features in ordinary telephone conversations over a cellular network or reproduce written text without calls in principle, but if this system has weaknesses, attackers will eventually identify them and be able to use them in advanced and extremely realistic schemes. social engineering.

Last week we already wrote about precautions to avoid becoming a victim of phone scams with fake voice. We recommend that you re-read the information, as well as remain vigilant even in communication with loved ones. Modern technology is much more insidious than it seems.



Source link

www.securitylab.ru

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Most Popular