A fake artificial intelligence (AI) generated video featuring popular SABC news anchors promoting an investment scheme has thrown the spotlight on sophisticated scams doing the rounds on social media and users have been warned not to believe everything they see.

The series of deepfake videos claim that individuals can earn up to R30 000 a day by investing in the project and tries to fool audiences into believing that the investment project has the endorsement of the SABC presenters and billionaire Elon Musk.

In another alarming case, the voice of SABC TV news anchor, Bongiwe Zwane, has been manipulated in an automated telephone message, soliciting funds for a foundation.

ALSO READ | Concerns over education system’s ability to handle issues of AI misuse

Both Zwane and the SABC have dismissed any association with the fraudulent activities and issued a warning to the public to be wary of digital scams that manipulate information.

Deepfake AI is a technology that employs deep-learning algorithms to create highly convincing fake images, audio and video content. The technology can replicate identities and mimic the actual voices to create ‘deepfake’ videos and sound clips.

While AI promises ground-breaking advancements in technology, healthcare and various other fields, anxieties persist regarding ethical dilemmas, job displacement, privacy invasion and the potential for independent systems to surpass human control.

As society grapples with the rapid evolution of AI, there is a growing consensus on the need for responsible development, transparent algorithms, and ethical frameworks to guide its deployment.

AI ethics and technology policy expert at the Africa Union, Lavina Ramkisoon, has advised South Africans to be vigilant and discerning consumers of online content.

In determining if a video is fake or AI-generated, Ramkisoon said people could look out for the tone and the image.

“Really look out for the voice in a message and pay attention to whether it is authentic. If you get an initial feeling that something’s not right, simply just don’t engage.

Secondly, look to tell if the images appear still-like, if that is the case, then that’s generally a good indication that it is AI-generated.

ALSO READ | Man and machine collaborate to make music

Ramkisoon also warned people to not fall for fake links.

“Don’t just click on something. If you receive an email, make sure that the email address that you receive it from is legitimate. You’ve got to take back that control and make sure that the information you’re sharing on any digital platform is legit.

“You can never be overly cautious,” she said.

Seth Bloomfield, a cybersecurity expert, said as deceptive content infiltrates our digital landscape, it is imperative for the government to pioneer comprehensive policies that safeguard the public.

“The unchecked proliferation of fake AI content threatens not only individual trust but the very fabric of our information ecosystem. Now, more than ever, we need proactive measures to mitigate the risks and preserve the integrity of our digital society.”