Deepfake voices deceive people – but not the brain

Deepfake technology can be used for fraud attempts on the telephone. Keystone / Gaetan Bally

The brain reacts differently to artificially imitated voices than to real ones. This happens even if people do not recognise the so-called deepfake voices as fake, according to a study by the University of Zurich in the journal Communications Biology.

Check out our selection of newsletters. Subscribe here.

Deepfake technologies can imitate a person’s unique voice profile very accurately. This is used, for example, for fraud attempts on the telephone, the university wrote in a press release on Wednesday.

For the study, the researchers recorded the voices of four male speakers and converted them into deepfake voices using computer algorithms. Twenty-five test subjects had to decide whether two voices they heard were identical or whether one of them was a deepfake voice.

In two-thirds of cases, the deepfake identities were correctly assigned. “This illustrates that although actual deepfake voices do not perfectly mimic identity, they have the potential to deceive people’s perceptions,” first author Claudia Roswandowitz was quoted as saying in the press release.

Brain areas react differently

However, the brains of the test subjects showed a different picture: the so-called nucleus accumbens, a component of the brain’s reward system, was much more active when both voices shown to the test subjects were natural voices. In contrast, the auditory cortex, which is responsible for analysing sounds, was more active when one of the two voices was a deepfake voice.

Roswandowitz concluded that humans can therefore only be partially deceived by deepfakes.

Translated from German by DeepL/ts

How we work

This news story has been written and carefully fact-checked by an external editorial team. At SWI swissinfo.ch we select the most relevant news for an international audience and use automatic translation tools such as DeepL to translate it into English. Providing you with automatically translated news gives us the time to write more in-depth articles.

If you want to know more about how we work, have a look here, and if you have feedback on this news story please write to [e-mail].