Deepfake Tech Used to Create This Call Between Swati Maliwal & Dhruv Rathee
The audio in this video has been created using a deepfake technology.
advertisement
A recording of a telephonic conversation purportedly between Aam Aadmi Party (AAP) leader Swati Maliwal and YouTuber Dhruv Rathee is going viral on social media.
What can he heard?: In the viral clip, Maliwal explains to Rathee how she was assaulted in front of Delhi Chief Minister Arvind Kejriwal and his wife Sunita Kejriwal.
She further requests Rathee not to make any video on this to defend them.
Further, Rathee also tells her that he gets paid by directly from the UK.
Background: Maliwal registered a complaint to Delhi Police and alleged that she was attacked by Kejriwal's aide, Bibhav Kumar on 13 May, inside the CM's residence.
Kumar has, however, denied the allegation. Following this, the Delhi police arrested Kumar on 18 May, and he is currently in the police custody.
Rathee's role: Rathee made a video on the incident and raised questions on the authenticity of this complaint.
How did we find out the truth?: There was no mention of this conversation of Maliwal and Rathee's social media profiles.
No report about this was covered by the media either.
We also noticed that at the 0:09 timestamp, there is a jump-cut and then an overlay of two voices/sentences.
AI detection tool: After examining the audio using AI detection tools, the Misinformation Combat Alliance’s (MCA) Deepfakes Analysis Unit (DAU) concluded that the audio is synthetic.
The voice detection tool of Loccus.ai indicated that this audio has more probability of being fake.
Screenshot of the analysis from Loccus.ai's audio detection tool.
(Source: DAU)
DAU also ran the audio through TrueMedia’s deepfake detector which overall categorised the audio as “highly suspicious”.
In a further breakdown of the analysis, the tool gave a 100 percent confidence score to the subcategory of “audio analysis” and a 99 percent confidence score to “advanced foundational features” — both indicators that it is highly probable that an AI audio generator was used to synthesise this audio.
The low confidence score of nine percent to the “AI generated audio detection” category suggests that the tool has more confidence in the audio having been synthesised using AI, and that it is not fully AI-generated.
Screenshot of the audio analysis from TrueMedia's deepfake detection tool.
(Source: DAU)
DAU also used the AI speech classifier of ElevenLabs, a company specialising in voice AI research and deployment. The results indicated that there was a very high probability that the audio file was generated using their software.
Screenshot of the analysis from A.I. speech classifier of ElevenLabs.
(Source: DAU)
DAU even shared the audio file for analysis with their detection partner ConTrails AI, a Bangalore-based startup, that has its own AI tools for detection of audio and video spoofs.
Their assessment suggested that they picked up patterns of AI voice cloning from both the voices heard in the audio.
Screenshot of the analysis from ConTrails AI's detection tool.
(Source: DAU)
They also mentioned that it is possible that techniques such as Retrieval-based Voice Conversion (RVC) cloning or Text-to-Speech (TTS) cloning were used to generate these voices.
Conclusion: Several deepfake detection tools confirmed that the viral clip allegedly between Swati Maliwal and Dhruv Rathee has been generated using AI.
(Not convinced of a post or information you came across online and want it verified? Send us the details on WhatsApp at 9540511818 , or e-mail it to us at webqoof@thequint.com and we'll fact-check it for you. You can also read all our fact-checked stories here.)
(At The Quint, we are answerable only to our audience. Play an active role in shaping our journalism by becoming a member. Because the truth is worth it.)