.jpg)
The Deepfakes Analysis Unit (DAU) analysed a video that apparently shows Rajat Sharma, India TV news anchor and chairperson, and Anant Ambani, non-executive director of Reliance Industries and their philanthropic arm, promoting a supposed new gaming app called “Aviator by Ambani”, which promises high monetary returns. After putting the video through A.I. detection tools and getting our expert partners to weigh in, we were able to conclude that synthetic audio was used to fabricate the video.
The video in Hindi, embedded in a Facebook post, was sent to the DAU tipline for assessment multiple times. The 10-minutes-and-19-seconds long video has only 42-seconds of audio-visual content followed by a series of static scenic images. It was uploaded on March 31, 2025 from an account with the display name of “Aviator 1” and since then it has garnered more than 270,000 views.
The display picture with the account is an illustration of a primitive aircraft. The profile information suggests that it is a “personal blog” and includes a link to some fraudulent gaming website, which asks for payment details to register; it does not mention the supposed app being peddled through the video. The other content posted from the account promotes some online gaming app named “Aviator” as a means to make big money. We don’t have any evidence to suggest whether this suspicious video originated from this account or another one.
The video has been packaged like a television news story with background music. The first segment opens with Mr. Sharma seated in a studio-like setting, with a colourful backdrop, apparently talking to the camera. Bold, static text graphics in Hindi can be seen at the top and bottom of the video frame throughout the video. The text at the top translates to “Be the next winner”. The bottom text translates to "Ambani's app: a success story that shook India”.
A male voice recorded over Sharma’s video track declares that an application called “Aviator by Ambani” can make someone a millionaire. It claims that the purported application has a life-altering potential and directs the viewer to supposed interviews with those who have apparently made money from the app.
The framing of the video shifts from a medium close-up of Sharma to an outdoor setting. A man with his back toward the camera can be seen holding a microphone with the words “India TV” emblazoned on it. He holds out the microphone to a man and woman who are part of a small group standing in front of him.
A male voice recorded over the visuals makes it seem as if he is asking each of them: “how much did you earn?”. A female voice recorded with the woman’s visuals replies “200,000” and a male voice, distinct from that of Sharma or the supposed interviewer, says “350,000 rupees”. Though the replies are fleeting, it is still evident that their lip-sync is imperfect and their teeth have an unnatural shine, which could be a sign of manipulation.
The segment that follows shows Mr. Ambani apparently addressing a press conference with several microphones bearing logos of media houses placed in front of him. The same clip and the accompanying audio was also part of a manipulated video that the DAU debunked through this report published last month. Synthetic audio was used with original footage featuring Ambani to fabricate the video.
Unlike this video, where the male voice with Ambani’s video track states clearly that a sum of “50,000 rupees” can be made through the supposed gaming app every week, in the video we previously debunked the voice did not clearly pronounce a number. Another distinction is that in the other manipulated video the purported app was referred to as “AVI by Ambani”. The media house logos visible in that video are not a part of this video and the Facebook account that posted the two videos are not the same either.
As mentioned in our last investigation, using a reverse image search, we were able to trace Ambani’s clip to this video published on Feb. 26, 2024 from the official YouTube channel of Business Today, a business news outlet. He is speaking in English in this video while the audio track of his manipulated segment is in Hindi.
In the last segment of the video, that we are analysing through this report, the same man holding the India TV microphone reappears, this time he is facing the camera with a water-body visible behind him. The male voice heard in this segment sounds the same as that in his earlier segment. That voice claims that the supposed app is a “gift for Indian citizens” and that people should download it and start making money right away. The video content stops abruptly and still images play out till the end.
The lip movements of Sharma, Ambani, and the man seen in the last segment align with the audio track. However, Sharma’s upper set of teeth is not visible in his segment at all and his lower set looks damaged with only a few teeth visible. Ambani’s teeth appear slightly blurred. The upper set of teeth of the man with the microphone is not visible at all and his lower set looks like a bright white patch, lacking definition.
The voice attributed to Sharma and Ambani, respectively, bears some resemblance to their real voice as heard in their recorded videos available online. However, their delivery sounds scripted and robotic, lacking pauses or intonation characteristic of human speech.
The audio of the man with the microphone also has a certain synthetic quality to it. As we did not know his identity or that of the supposed interviewees we could not use a keyword search to find any videos online, for comparison.
We undertook a reverse image search using screenshots from the doctored video. The segment featuring the man with the microphone, and the clips of the supposed interviewees were traced to this video published on Jan. 11, 2025 from the official YouTube channel of India TV, a Hindi news channel led by Sharma. And this video also published from the same YouTube channel on Feb. 10, 2025 carried the original footage featuring Sharma.
The clothes, backdrop, and body language of the speakers in their respective segments in the doctored video and the videos we traced are identical. However, more zoomed-in frames have been used in the doctored video. The background music in the doctored video is not part of any of the original videos we traced.
The India TV logo is prominently visible in both the original videos. In Sharma’s original video, which is in Hindi, there is no mention of any gaming app. It also carries a news ticker at the bottom of the video frame along with text graphics both at the bottom and top of the frame.
We discovered that the man with the microphone is a reporter with India TV. The other man and woman were actually interviewed for a news segment featuring the reporter at the Maha Kumbh, a Hindu religious congregation organised earlier this year in India’s Uttar Pradesh state. Neither of the three mention any gaming app or a sum of money. The reporter and the woman converse in Hindi while the conversation between the man and reporter is recorded in English.
The reporter speaks in Hindi while standing before a water-body in the original video, a portion of this visual has been lifted and used in the doctored video but of course without the same audio. The reporter’s delivery sounds lively and animated in the original video compared to that in the doctored video.
This is not the first time that original footage featuring Sharma has been manipulated using synthetic audio to fabricate videos. In December 2024, he clarified in a post on X that his cloned voice is being misused to promote fake remedies. We have debunked several such videos through our reports such as this, this, and this.
To discern the extent of A.I. manipulation in the video under review, we put it through A.I. detection tools.
The voice tool of Hiya, a company that specialises in artificial intelligence solutions for voice safety, indicated that there is a 27 percent probability of the audio track in the video having been generated or modified using A.I.
Hive AI’s deepfake video detection tool pointed to a small segment in the video track, featuring the supposed interviewees, as being A.I.-manipulated. The tool did not detect manipulation in the rest of the video. Their audio detection tool, however, indicated A.I. manipulation in most of the audio track but for a 10-second segment purported to be Sharma’s voice.
We also ran the audio track from the video through Deepfake-O-Meter, an open platform developed by Media Forensics Lab (MDFL) at UB for detection of A.I.-generated image, video, and audio. The tool provides a selection of classifiers that can be used to analyse media files.
We chose seven audio detectors, out of which four gave strong indicators of A.I. manipulation in the audio. AASIST (2021) and RawNet2 (2021) are designed to detect audio impersonations, voice clones, replay attacks, and other forms of audio spoofs. The Linear Frequency Cepstral Coefficient (LFCC) - Light Convolutional Neural Network (LCNN) 2021 model classifies genuine versus synthetic speech to detect audio deepfakes.
RawNet3 (2023) allows for nuanced detection of synthetic audio while RawNet2-Vocoder (2023) and RawNet2-Vocoder-V1 (2023) are useful in identifying synthesised speech. Whisper (2023) is designed to analyse synthetic human voices.
For a further analysis on the audio track we also ran it through the A.I. speech classifier of ElevenLabs, a company specialising in voice A.I. research and deployment. The results that returned indicated that it was “very unlikely” that the audio track in the video was generated using their platform.
For expert analysis, we escalated the video to our detection partner ConTrailsAI, a Bangalore-based startup with its own A.I. tools for detection of audio and video spoofs. The team ran the video through audio and video detection models; and concluded that the audio and video appear to be generated or manipulated using A.I.
In their report, the team noted that their model predicted that the audio of each speaker is A.I.-generated. They added that the pacing is unnatural in certain parts and relatively monotonous compared to real speech.
They stated that the video shows signs of A.I. manipulation as the lip movements appear animated synced with the A.I.-generated speech. In the frame-by-frame analysis shared below, they have first analysed Ambani's segment followed by Sharma’s and then the reporter’s.
To get another expert analysis on the video, we escalated it to the Global Deepfake Detection System (GODDS), a detection system set up by Northwestern University’s Security & AI Lab (NSAIL). The video track was put through seven deepfake detection algorithms, 20 deepfake detection algorithms were used to analyse the audio track; and the video was overall analysed by two human analysts.
Six of the seven predictive models gave a lower probability of the video being fake and one indicated a high probability of the video being fake. Eleven of the 20 predictive models gave a high probability to the audio being fake while nine gave a lower probability to the audio being fake.
The team highlighted several instances of A.I. manipulation in the video such as Sharma’s teeth appear to be missing, which we too had pointed to above; and Ambani’s mouth, nose changing shape. They further noted that the last second of the video features a clip of the sky, which does not match the video's news-like content and formatting.
In their overall verdict, the GODDS team concluded that the video is likely fake and possibly generated using artificial intelligence. They added that their analysis suggests the background music present in the audio channel may possibly have been added as an anti-detection measure.
Based on our findings and analyses from experts, we can conclude that separate video tracks of Ambani and Sharma have been stitched together and used with synthetic audio to fabricate the video. This is yet another attempt to promote a dubious gaming app in a bid to scam the public.
(Written by Debraj Sarkar and Rahul Adhikari, edited by Pamposh Raina.)
Kindly Note: The manipulated audio/video files that we receive on our tipline are not embedded in our assessment reports because we do not intend to contribute to their virality.
%20(1).jpg)
