That day Rosiana Silalahi, Editor-in-Chief of Kompas TV, was in her office when a short message arrived on her cell phone. A friend asked her if the video circulating on social media was really her. She was quite surprised and hurriedly checked her handphone. There she found a video showing her with Dr. Terawan. In the video they talked about methods of curing hypertension.
Rosi immediately fainted. The reason is that the interview really happened, the event really happened. She remembers very well when she came to Dr. Terawan’s house in Central Jakarta for an interview. But in the video sent by her friend, her voice was changed. The sentence she said was different from the conversation that took place with Dr. Terawan at that time.
For a moment, her face became tense when she checked the comments on the video. Moreover, the same message was also sent by several people, including by her colleague, a doctorate in economics.
“Is this true or a hoax, Rosi?” was the question that came into her cell phone.
Recently, videos have appeared on social media featuring female journalists interviewing experts or public figures. The video discusses a disease that many people complain about or a site that can bring in money.
One of them is a video featuring Rosi interviewing Dr. Terawan who discusses methods of curing hypertension. Checks from a number of media fact-checking teams show that the videos circulating on social media are hoaxes or false information.
“So it’s like a coin between being angry because people can really sell with fraud. But it also makes me sad because I read in the comments column that there are people who have already bought for their sick father,” Rosi told Konde.co, Wednesday (18/9/24).
Read More: America’s Glass Ceiling Remains: Here are Some of the Reasons Why a Woman May Have Once Again Lost the Presidency
Rosi does not blame people who are fooled by the video that has been manipulated. Moreover, the person who speaks is an authority in her field, namely a doctor.
“I can’t immediately judge them, ‘How come you believe it’. Because if I see here (there is) Mr. Terawan and he is famous as a doctor who does have his own way of healing patients. So, I feel sorry for those who have sick relatives and are consumed by this lie,” said Rosi.
From the videos manipulated using audio-based generative artificial intelligence technology, there are similarities that can be identified. One of them is that content creators use talk show videos hosted by female journalists or news programs with female presenters. This means that content creators use journalistic products to spread false information.
It’s not just once that Rosi has experienced this. One of the latest videos containing disinformation also uses Rosiana’s video. The video circulated on social media in early August 2024. A Facebook post the video and netizen immediately talk about it. The video is a snippet from the ROSI talk show program that airs on KompasTV with Rosianna Silalahi as the host.
Read More: How The Playboy Bunny Suit Went from Uniform to Halloween Costume
Apart from Rosi, the video also featured a cardiologist, Isman Firdaus. The video discusses a home-made ginger drink that normalizes blood pressure. In the video, Rosi and Isman look real and speak in voices that sound similar to theirs.
“Good news, the problem of high blood pressure in Indonesia has ended forever. Cardiologist Isman Firdaus revealed the dirtiest secrets of pharmaceutical companies that they hid from everyone for 77 years,” said Rosi, opening the conversation in a flat tone.
If we really listen to the video, there are a few seconds where Rosi’s voice slips and there are repetitions of words in the numbers section that are not natural. So it sounds like the voice of a bot or artificial intelligence (AI).
“Doctor Isman Firdaus discovered the truth and found out about a very simple drink that normalizes blood pressure to 120/80 in just 27 hours of use. Now Doctor Isman himself will tell us about this simple home brew,” Rosi’s voice continued.
Read More: The Terminator at 40: James Cameron’s Dark Vision is More Relevant Than Ever
Isman then responded to Rosi’s words. He spoke with a different video background from Rosi’s.
“My name is doctor Isman Firdaus, I have been studying hypertension and high blood pressure for 25 years and have become one of the leading cardiologists in Indonesia,” he said.
Isman revealed that he had conducted hundreds of scientific experiments and concluded that there was no drug that could stabilize a person’s blood pressure. He said it in a flat tone, almost the same from one sentence to another.
He offers a solution in the form of a drink from ginger root that is claimed to be able to lower the blood pressure of thousands of his patients. Isman also claims that the drink has been certified and tested through clinical science. He advises viewers to watch a video tutorial on how to make and use the drink. But the video will be deleted in 15 minutes for his own safety.
Since the Facebook account was uploaded on August 4, 2024, Rosiana’s video has been viewed 3 million times and liked by 19 thousand people. This video was also shared by 1,800 Facebook users and commented on by 975 netizens.
Read More: How Should You be a Young Person in A Time of 26USD Buns and Palace Family of Nepotism?
Most of the comments from the netizens were thanks to the doctor. This kind of comment usually gets a lot of positive emoticon reactions from netizens. There were also some comments from netizens who regretted that the doctor ended up selling medicine. Only a small portion of the comments showed that they knew the video was AI. Some of those who knew also warned other netizens not to be fooled by hoaxes.
Not only Rosiana, the presenter of Narasi TV, Najwa Shihab has also experienced. Like the video uploaded by an Instagram account on January 15, 2024, the video shows an interview between journalist Najwa Shihab and two celebrities, Raffi Ahmad and Atta Halilintar. The interview discussed the Kobe138 site, an online gambling site that offers slot games where players can place bets online. In the video Raffi and Atta said that each player would benefit. This is because they both have disbursed a number of funds to invest in the site.
The cases above show that deepfake videos uploaded on social media are used for marketing purposes. Whether it is offering and selling health products or inviting people to join online gambling, etc.
Fact Check Tracking
Social media coverage data of the content shows that Rosiana’s nearly three-minute video has spread widely in a short time. This data is just from one account on one social media. It is possible that other accounts also spread it on other social media platforms. This is because social media users have a tendency to share content across platforms, especially on viral content.
This video is one of a number of similar videos made using deepfake technology.
Tempo’s fact-checking showed that the video came from two videos made for two different events. The time when the videos were made is also not the same.
The video part of Rosi was taken from the talk show program ROSI. In the original video, Rosi presented two Indonesian men’s singles badminton players, Jonatan Christie and Anthony Sinisuka Ginting, who became finalists in the 2024 All England Championships tournament in Birmingham, England. The video aired on KompasTV’s YouTube channel on March 21, 2024.
The Rosi video piece was then combined with the Ramadhan Tausiyah video which presented Isman as a guest speaker. Isman was the speaker in episode 10 which discussed “Tips to Maintain Heart Health ‘035 140 530’ during fasting and Eid. The video was aired on the YouTube channel of the Islamic Studies Forum of the Faculty of Medicine, University of Indonesia on May 12, 2021.
Read More: ‘Black September’, Diva Suukyi’s Story and the Victims of Human Rights Violation Cases
Neither Rosi nor Isman were talking about ginger root drinks to lower blood pressure. They were talking about two different topics. However, Rosi and Isman’s voices were changed from the original video using artificial intelligence (AI) technology. The findings of Tempo’s fact-check team stated that Deepware analysis showed the use of Voice Generator AI between 97-99 percent.
As for the video of Rosi interviewing Terawan, the Kompas.com fact- check team explained that the video uploaded by the Facebook account was taken from Rosi’s interview with Terawan for the ROSI program. In the original video, Rosi and Terawan discussed a number of public questions such as the controversy over the brainwashing therapy method and the archipelago vaccine. This interview was uploaded on KOMPASTV’s YouTube channel on July 8, 2022 titled “[ROSI EXCLUSIVE] Finally, Doctor Terawan Answers”.
The results of the fact check explained that the original video sound was changed using audio-based generative AI technology. Tests conducted with an AI voice detector device found that Rosi’s voice was detected as AI-made with a probability of 79.64 percent and Terawan’s voice 79.60 percent.
In the video, tvOne presenter Agita Mahlika discusses a diabetes drug made by an Indonesian doctor that can cure diabetes in three days. Furthermore, the former Minister of Health, Terawan Agus Putranto, promotes the drug and claims 20 thousand people have tried it and recovered from diabetes. The video also contains testimonials from a number of people who claim to have suffered from diabetes but now their blood sugar levels have dropped.
Read More: ‘Black September’, Diva Suukyi’s Story and the Victims of Human Rights Violation Cases
Content creators create deepfake videos by combining a number of videos with different contexts. In this video, the voices of the people appearing in the video are not as “natural” as the videos made in 2024.
Deepfake videos circulating on social media not only contain disinformation, but also scams or fraud related to online gambling. This kind of video circulated in January 2024 using the figures of female journalists and celebrities who have millions of followers.
Therefore, according to Rosi, content creators use women as tools, or tools to sell. This is inseparable from the stereotypes prevailing in society that assume that women are more capable of selling than men.
On the other hand, it can be seen as a blessing in disguise, showing that women are much more trustworthy than men. Even so, this situation is also prone to placing women as objects, whose bodies are exploited. According to Rosi, this is a patriarchal and demeaning way of selling to women.
When the Rosi deepfake video spread on social media, KompasTV‘s legal team then wrote to the platform company, they made a report and complaint. This effort could curb the spread of deepfake videos. Even so, this step is not enough.
Read More: ‘Agatha All Along’ : The Marvel Comics Witch Who Broke Societal Barriers, Why?
After the report, the deepfake video of hypertension medication uploaded in January 2024 is currently inaccessible on the Facebook account that uploaded it. Meanwhile, the video uploaded in August 2024 can still be viewed, commented on and shared.
Rosi said efforts to report have been made, but the possibility of similar videos appearing remains. Moreover, it becomes a kind of advertisement (ads) that can appear at any time.
“Yes, at that time we reported it on YouTube. But like you, yesterday (the video) still appears on YouTube. So there is still (the video) even though we have reported it. And it became part of the advertisement,” he said.
The Dangers of Deepfake Disinformation; AI Risks for Journalists and the Media
It can be said that when one video is successfully deleted, other new videos will appear. This situation is supported by the easier use of AI technology. Moreover, AI technology is now freely available, as explained by Ika Ningtyas, a journalist at Tempo Fact Check.
“The deepfake technology that is now freely available is actually low cost and more accessible. It also doesn’t really require technology skills. For example, people used to have to understand Photoshop or Corel Draw. But with generative AI technology, now people only need to give commands or prompts to AI. So it is that easy now to create content that can be faked,” Ika Ningtyas told Konde.co, Sunday (15/9/24).
Meanwhile, the results of these deepfake videos are very similar to the original content. From the deepfake videos that have been circulating, for example, the 2024 video is more “smooth” than the videos produced in previous years. The voice produced is so similar to the voice of the original speaker that it is increasingly difficult for people to distinguish between original content and AI content. Consequently, false or misleading information spread by using deepfake videos can increasingly fill our digital public space.
Read More: The Madness toward Major Teddy on TikTok: Military Idolization on Social Media
Moreover, the materials to create them are abundantly available on social media, such as text, photos, videos, etc. This condition will affect the lives of the wider community, for example related to public health issues, politics, democracy, and so on. This condition will affect the lives of the wider community, for example related to public health issues, politics, democracy, and so on.
“For example, Najwa or Rosi are like selling medicine. People who don’t know that this is a fake video will definitely buy the medicine. In the future, there will be more deepfakes like this that can endanger public health,” said Ika Ningtyas.
“This can also endanger our democracy, including triggering violence if. For example, deepfakes are used to provoke or increase hatred towards minority groups,” she added.
Coping Strategies for Deepfakes: Distinguish Fact from Fiction
Arif Perdana, Associate Professor of Digital Strategy and Data Science, Monash University, in his article on Theconversation.com stated that advances in generative AI algorithms, and the ability to generate synthetic media are no longer monopolized by film and video editing professionals. Ordinary people can access this technology to produce complex content quickly, easily and at low cost.
One of the problems is that misused AI versions of synthetic media can have serious consequences. For example, when used to create “deepfakes,” which is synthetic media content created with the intention of misleading or committing a crime. Deepfakes can trigger panic and confusion in an already tense situation.
This incident shows how sophisticated technology can be misused to deceive and cause great material harm, from privacy, defamation, copyright infringement, financial loss, to social unrest.
Arif Perdana provides steps so that we are not fooled by deepfakes. First, we need to recognize inconsistencies in the face, hair and elements in the video. The second is to use AI algorithms to test for these inconsistencies. The third is to change it by involving legal policies to address it. Fourth, it increases public awareness to know what is real and what is fake. And finally, platforms play a crucial role in maintaining the integrity of information by providing verified and reliable sources.
Read More: Stereotypes of Javanese Women in Films: Roles as Domestic Worker, Poor, and Plebeian
Ika Ningtyas added that journalism has a role to oversee the development of AI technology. So journalists not only use but must oversee the development of AI technology. This supervision is especially against its adverse effects or misuse that could risk human rights.
“We must proactively monitor how AI technology can negatively impact human rights,” Ika explained.
This situation is certainly a challenge for the work of journalists and mass media. Ika sees that journalists and the media actually have a big role when false information using AI technology is increasingly abundant.
“Journalists are even more relevant because they will be the frontliner to verify, help the public provide correct information on existing content. So that they know which content is real and which is not.”
That way AI technology helps journalism to detect fake content. Therefore, according to Ika, the adaptation of AI in the field of journalism is important. Journalists need to master the skills to dismantle deepfakes and have the proficiency to use technology to detect AI.
Overseeing here means keeping an eye on the accountability of AI tech companies so that they don’t lose sight of their accountability aspects. It also means keeping an eye on how AI technology can negatively impact the most vulnerable groups. This is because AI technology is widely used in security, transportation and so on.
Read More: Love Languages Are Hugely Popular. Are They Real?
Another important thing is to ensure that the state can comprehensively regulate AI technology. Comprehensive means that the regulation does not kill innovation related to the creation of AI, but on the other hand, the AI technology created must be in line with the principles of human rights.
According to Ika, journalists can take a role in ensuring how transparency, ethics, protection of innovation, complaint handling procedures are regulated by the state. In short, journalists and the media oversee it so that everything can run.
This content was created by Konde.co with funding from the Indonesia Media Program, based on Konde.co’s editorial policies and guidelines while maintaining Konde.co’s editorial independence.
This article is part of the #StopNgelesUUTPKS Series, a series on coverage of sexual violence cases and the implementation of The Law on Sexual Violence Crime (TPKS) Law.
Translator: Theresia Pratiwi Elingsetyo Sanubari
