With the help of new voice artificial intelligence, it will be possible to hack bank accounts

With the help of new voice artificial intelligence, it will be possible to hack bank accounts

A tech journalist recently discovered that AI voice generation software can be used to breach bank account security measures. The software was used to replicate a synthetic voice that could access account information, including balances and recent transactions. This article will discuss the dangers of AI voice generation software, how it is used, and what measures banks can take to prevent fraud.

The dangers of AI voice generation software

AI voice generation software can create synthetic voices that can speak any text entered into a website. This puts everyone at risk, as anyone with access to a voice recording could potentially use software to replicate someone’s voice and gain access to sensitive information.

Joseph Cox, a tech writer for Vice, recently tested his bank’s automated security measures using a recording he made using an AI voice generator. Within seconds, he had access to his account information, including balance and recent transactions. He later tried the same thing with a Lloyds bank account, where his first attempt was unsuccessful, but after reading longer texts to the software to improve the credibility of the accent, he still managed to gain access.

How does AI voice generation software work?

To use AI voice generation software, a person records a few minutes of speech and uploads it to the software. The software then creates a synthetic voice that can speak any text entered on the page. The software is easy to use and does not have strong protection mechanisms to prevent misuse.

Banks use voice authentication technology

Many banks, including Wells Fargo, TD Bank and Chase, use voice authentication technology to allow customers to perform banking activities over the phone, such as checking account balances and transaction history and transferring money. Lloyds Bank boasts that its Voice ID program is secure and unique to each person, analyzing more than 100 characteristics, but it can’t seem to distinguish between human voices and those generated by artificial intelligence.

Possible consequences

As AI tools continue to develop and improve at a rapid pace, what now appears to be the ultimate authentication measure could soon become very easy to crack. Rachel Tobac, CEO of SocialProofSecurity, recommends that organizations using voice authentication switch to a secure method of identity verification, such as multi-factor authentication, as soon as possible. AI voice generation technology allows people to hack someone’s account without ever interacting with that person in real life.

Podijeli članak:

Facebook
Twitter
Reddit
WhatsApp