Deepfake ads with celebrities raise red flags, Govt says risky

[ad_1]

Shahrukh Khan appears to be ostensibly advertising a betting platform on Instagram, claiming that “everyone wins in the game”. So does Virat Kohli, who seems to bat for a betting app with claims of big gains that he’s personally made.

But, there’s a catch. Their audio and video are slightly out of sync.

That is because while it appears that Khan and Kohli are making these claims on Instagram and Facebook advertisements, which have been viewed tens of thousands of times, it is not actually them – but digitally altered versions of their face and voice to make it seem like them.

Deepfakes of popular celebrities have arrived on the scene to hard sell products and services to impressionable fans, especially contentious products replete with lofty claims.

On Instagram and Facebook, which collectively have hundreds of millions of users in India, advertisers are marrying the worst of two worlds to ostensibly deceive people: deepfakes of prominent Indian celebrities from across entertainment and sports who are made to appear to be promoting offshore betting platforms with dubious claims.

A deepfake is a video of a person in which the face or body has been digitally altered so that he or she appears to be someone else, typically used to spread false information.

Prime Minister Narendra Modi recently raised concerns about the potential harms of deepfakes, and the government has initiated a crackdown on such content on social media platforms. On Thursday, the Ministry of Electronics and Information Technology (MeitY) is meeting with representatives of social media platforms to discuss the issue, and a standard operating procedure is expected to be shared with them on how to deal with such content.

Instagram and Facebook – which are owned by Meta Platforms – are rife with advertisements and sponsored posts which contain deepfakes of people, including Shahrukh Khan, Virat Kohli, Akshay Kumar, and a number of popular digital creators, as per an analysis by The Indian Express.

Explained

The threat, the alert

In one of the videos, a game called Aviator was being promoted through a deepfake of Shahrukh Khan. In the game, people bet on a virtual plane flying on the screen and have to cash out before it crashes out. The deepfake video of Khan was found making dubious assertions in the ads promoting the Aviator game, including claims like “everyone wins in this game”. This video had over 50,000 views.

In the video with Virat Kohli’s deepfake, a newscaster from an Indian TV channel is seen reading a headline about how he has won Rs 8,00,000 on a betting game, followed by a digitally altered Kohli saying that he has won “a huge amount of cash” playing the game and how his “hands are still shaking” after the win.

Queries sent to representatives of Khan, Kohli, and Kumar did not elicit an immediate request for comment.

In response to a detailed questionnaire on The Indian Express’s findings, a Meta spokesperson said the company removes content when it is made aware of it.

“We remove content when we are made aware of it and when it violates our policies whether it was created by AI or a person. Advertisers running ads across Meta technologies must follow our Community Standards and our Advertising Standards. We use automated and, in some instances, manual review to enforce our policies,” the spokesperson said.

Such ads are paid for by advertisers on Instagram and Facebook and specifically targeted at users from the Indian region, meaning that Meta is likely making money out of hosting artificially manipulated content, at least for the duration that they are live on its platforms. Digital advertising is the backbone of Meta’s business model globally and in India.

As per Meta’s blog on how it reviews ads, the company says that reviewing ads from millions of advertisers globally against its advertising standards is essential, but it is not without challenges. “Our enforcement isn’t perfect, and both machines and people make mistakes,” the company says on its site.

However, the company does say that ads may be reviewed again at any point, including after they are live.

Facebook India Online Services, the flagship registered entity for Meta in the country, clocked gross advertisement revenues of Rs 16,189 crore in fiscal year 2021-22, a 74 per cent year-on-year growth. Facebook India purchases advertisement space from Meta Platforms, its parent company in the US. Meta India applies some markup to the cost of goods sold when selling goods to their Indian customers.

These dubious advertisements spotlight the uphill battle that social media companies face in weeding out such content from their platforms. A problem only compounded by the ease with which people can today create deepfakes.

Mostly, a controversy erupted after a video of actor Rashmika Mandanna entering an elevator went viral on social media. What initially appeared genuine was, in fact, a deepfake of Mandana. The original video featured a British Indian person, and their face was morphed to insert Mandanna’s face instead. Not only this, there has been a surge of deepfake audio and videos of political leaders on platforms like Instagram.

Earlier this month, MeitY had sent advisories to social media platforms, including Facebook, Instagram and YouTube, to take down deepfakes. The advisory had reiterated existing legal provisions that platforms have to follow as online intermediaries. It had mentioned Section 66D of the Information Technology Act, which entails punishment for cheating by personation by using computer resources with imprisonment of up to three years and a fine up to Rs 1 lakh.

The advisory also mentioned Rule 3(2)(b) of the Information Technology Rules, under which social media platforms are required to take down content in the nature of impersonation, including artificially morphed images of an individual.

[ad_2]

Source link