PM Lee expresses concern over ‘deepfake’ technology that can mimic people’s voices through AI

Singapore Prime Minister Lee Hsien Loong has expressed concern over ‘deepfake’ technology, as the controversial artificial intelligence-based technique draws criticism worldwide.

‘Deepfake’ is a human image synthesis technology that uses artificial intelligence to superimpose faces onto source images or videos or mimic the voices of individuals. These capabilities have given rise to cases of fake pornography and malicious hoaxes.

In a Facebook post published today (19 Sept), PM Lee expressed concern over this startling new technology that came into prominence in 2017. Referring to reports of deepfake artificial intelligence being used to scam companies through voice fraud, PM Lee said:

“Technology cuts both ways. While it has undoubtedly made our lives easier, it has also enabled new and insidious types of criminal activity – cyber crime. As Artificial Intelligence becomes more sophisticated, it is becoming more difficult to believe what you see or hear.”

Noting that technology has evolved where cybercriminals who used to create fake ads using his photo are now able to fake voices to trick victims, PM Lee referred to a case that was reported earlier this month where a voice deepfake was used to scam the CEO of a UK-based energy firm out of USD $243,000.

PM Lee said: “First, there were fake ads using my photo. Now, criminals are faking voices to trick their victims. In the first so-called ‘AI heist’, criminals used voice-imitating AI to mimic a CEO’s voice, to trick his MD into transferring several hundred thousand dollars of company funds to them.

“Such AI programs can be easily trained to mimic voices and speech patterns from public sources like interviews and YouTube videos. Some programs are available for free, so anyone can download and use deepfake software to recreate anyone’s voice.

“We have to adapt quickly. Tech companies are developing ways to detect deepfake audio and video. Meanwhile, we should be discerning about what we come across online, whether or not it seems legitimate. Always be careful. If in doubt, check first before you proceed.”