Wystarczy 15 sekund próbki głosu i oszustwo gotowe

Next generation fraud. There is no way to protect against this

Faking someone's voice is becoming easier and easier. OpenAI's latest tool requires just a 15-second audio sample.

Artificial intelligence opens up incredible opportunities for us, but at the same time it poses enormous threats. Today, the Internet is full of deepfakes. Even faking someone's voice takes only a moment for a knowledgeable person. As if that wasn't enough, it's getting easier.

New OpenAI tool

OpenAI, which you should know primarily from Chat-GPT, has introduced a new tool called Voice Engine. The scary thing is that it only requires a 15-second sample of someone's voice to successfully imitate it.

OpenAI assures that 15 seconds is enough to reproduce someone's voice. After adding the appropriate sample, just enter the text that you want to convert into speech and after a while the magic happens. Suddenly Messi can say he loves Real Madrid and Elon Musk can praise Google, Apple and Microsoft.

OpenAI does not provide its tool yet. The company is aware of the risks associated with such a simple voice counterfeiting mechanism. The company wants to start a dialogue on the responsible use of similar solutions.

Based on these conversations and the results of these small-scale tests, we will make a more informed decision about whether and how to implement this technology on a large scale.

– informs OpenAI.

It's not just about scams, which are already abundant on the Internet. OpenAI is aware that such a solution may also negatively affect the work of voice actors, who may turn out to be unnecessary.

Similar Posts