Deepfakes Explained: How AI-Generated Video Threatens Privacy and Identity

On 3 October 2026, actor Jr NTR said he would pursue legal action after an AI-manipulated video based on the song "Chuttamalle", from the 2024 Telugu film Devara: Part 1, began circulating online, according to a report by ThePrint. The clip digitally altered the appearance of co-star Janhvi Kapoor to create sexually explicit imagery, the report said.

The Chuttamalle video controversy

Jr NTR posted a statement on X describing the video as disgusting, sick and shameless, ThePrint reported. He said he would not spare those responsible and would take every possible legal action against the people who created the clip and those spreading it, adding that he wanted stronger government regulation of such AI misuse. Janhvi Kapoor separately urged the public not to create, share or promote the clip and said legal action would be taken against those involved, per the same report.

Actor Vijay Deverakonda publicly backed Jr NTR's call for stronger laws against AI misuse, according to AP7AM. He argued that producing this kind of manipulated content should be treated by the legal system with the same seriousness as sexual assault, and he called for the people responsible to be punished as a deterrent.

What a deepfake actually is

The Chuttamalle video is not an isolated case. In November 2023, a manipulated clip swapped actor Rashmika Mandanna's face onto the body of a different woman, British-Indian social media user Zara Patel, filmed in a lift, according to The Tribune. Mandanna said at the time that the experience was extremely scary, not only for her but for women in general, and warned that people without support systems could be especially vulnerable. Her Goodbye co-star Amitabh Bachchan responded that it was a strong case for legal action, the same report said.

A deepfake uses AI models trained on real images, video or audio of a person to generate new footage that makes it appear they said or did something they did not. The same underlying technology can clone a voice, swap a face onto another body, or generate an entirely synthetic scene, which is why both the Chuttamalle and Rashmika Mandanna incidents involved transplanting a celebrity's likeness onto someone else's recorded footage.

How widespread the problem has become

Data compiled by research firm DeepStrike and cited by StationX shows the volume of deepfakes circulating online rose from about 500,000 in 2023 to roughly 8 million by 2025. Separate figures from identity-verification firm Sensity AI, also cited by StationX, indicate that non-consensual sexual content makes up the large majority of deepfake videos found online, and that the overwhelming majority of people depicted in them are women.

India's new rules on synthetic media

India's Ministry of Electronics and Information Technology (MeitY) amended the IT (Intermediary Guidelines and Digital Media Ethics Code) Rules in an announcement dated 22 October 2025 to explicitly cover what it calls "synthetically generated information" — content that is artificially or algorithmically created, generated, modified or altered and appears reasonably authentic, according to a legal analysis by Khurana & Khurana.

• Visual deepfakes must carry a permanent on-screen label covering at least 10% of the frame for the full duration of the clip.
• Audio deepfakes must carry an audible disclosure during the first 10% of playback that cannot be stripped out by platforms or users.
• Platforms with more than 5 million monthly active users must deploy technical measures to verify whether uploaded content is synthetic.
• A separate court ruling in the Sadhguru case, from May 2025, set a 36-hour window for platforms to remove infringing deepfake content once notified.
• Private messages and unpublished drafts fall outside the scope of the new labeling and takedown duties.

The amendment does not by itself resolve cases like the Chuttamalle video, which both Jr NTR and Janhvi Kapoor say they will pursue through existing legal channels rather than the new labeling framework. But it marks India's first rule set written specifically around AI-generated impersonation, placing takedown and verification obligations directly on large platforms rather than leaving enforcement solely to individual lawsuits.