The Aam Aadmi Party (AAP)’s AI-generated posts showing Delhi Chief Minister (CM) Rekha Gupta as a liquor vendor are still on X, four days after the Delhi High Court (HC) ordered their removal. On November 6, the court directed that six posts be taken down “forthwith”. As of November 10, all five posts on X were still live. Only the post on Facebook has been taken down.
One of the videos has been viewed over 512,000 times. AAP Delhi posted it on November 3 with the caption: “Behind the scenes of ‘Har Har Rekha, Ghar Ghar Theka’ Yojana.. Absolute Chaos!” It appears to be the video Gupta’s lawyers described in court, which shows the CM and Delhi’s Deputy CM “together, pouring liquor and ice-cubes”. A post from November 4 shows her carrying a basket of liquor bottles on her head. It has over 72,000 views.
“In the garb of freedom of speech and expression and under the disguise of humour and satire, no one can be permitted to be reduced to an object of ridicule and insult,” Justice Manoj Jain said in the order.
Her lawyers told the court that the content depicts her as a “liquor vendor, hawker and delivery agent” and as the head of a “non-existing ‘Har Har Rekha or Ghar Ghar Theka’ Yojana”. The judge agreed that the content gives “an impression as if plaintiff herself is a liquor-vendor”.
These are the six posts the court ordered removed:
- AAP Delhi on X, November 4: Gupta carrying liquor bottles. Still live.
- AAP Delhi on X, November 3: “Har Har Rekha, Ghar Ghar Theka” video. Still live.
- AAP Delhi on X: Video. Still live.
- AAP Delhi on X: Post. Still live.
- AAP’s national account on X: Post. Still live.
- AAP on Facebook: Video. No longer available.
Why did it take a court order?
Justice Jain said courts have a duty to intervene until AI-generated content is properly regulated. He relied on an earlier Delhi HC order in a case filed by Bharatiya Janata Party (BJP) Member of Parliament (MP) Raghav Chadha:
“Till the time publication of such content is not regulated appropriately by a stringent legislation, it becomes judicial duty upon receipt of grievance to examine whether the use of AI has reached the threshold of infringing upon an individual’s fundamental rights of dignity or not.”
India does have rules governing such content. The Information Technology (IT) Amendment Rules 2026 on synthetically generated information have been in force since February 20. The order does not mention them. Under these rules:
- Labels: Platforms must clearly and prominently label AI-generated content.
- Declarations: Significant social media intermediaries (platforms with over 50 lakh users in India) must ask users to declare whether their content is AI-generated. They must also verify these declarations.
- Timelines: Platforms must comply with lawful takedown orders within three hours, down from 36 hours. They must also act on certain complaints within two hours, down from 24 hours.
However, AAP’s videos remained online for days before the CM approached the court, and they are still on X despite the order. The order does not say whether she first complained to X or Meta. This raises two questions:
- Implementation: How are the rules on synthetically generated information being implemented?
- The court route: If the rules are already in force, why did the CM have to approach the court to get the content removed?
Satire or misinformation?
The court also considered whether the videos amounted to fair political commentary. “There is always a thin line between a political satire and offensive depiction,” Justice Jain said. He found that AAP had crossed that line.
“Admittedly, any political party is entitled to criticize any governmental policy or decision or administrative action of the government but in the garb of expression of disapproval of any such policy, the offensive portrayal of Chief Minister, with the assistance of AI-generated videos, hurting her very dignity can be, by no stretch of imagination, said to be in the realm of permissible political satire,” he said.
MediaNama founder Nikhil Pahwa sees it differently. “What they have published is effectively satire,” he said.
“This is satire. Very clearly it cannot be taken as a realistic occurrence. And so, what is wrong if satire is used?” Pahwa said.
The videos do look real. Gupta’s lawyers argued that “the video-images look so real and near-perfect imitation”. However, what they depict is a scheme that does not exist. AAP also labelled at least one of the videos as AI-generated. The November 3 video carries a small “AI GENERATED” watermark in one corner.
The IT Rules apply to content that is likely to be perceived as “indistinguishable from a natural person or real-world event”. However, they do not specify how realistic content must be to meet this threshold.
“If this is realistic or deemed to be realistic, how realistic is real? Like, is it a 60% likelihood? Is it a 70% likelihood? Is it an 80% likelihood?” Pahwa asked. “Is it misinformation if it clearly cannot be construed as real?”
Do other parties do this too?
AAP is not the only political party using AI-generated videos against its rivals. In June, we identified 10 instances in which India’s deepfake rules were not enforced. Four of these involved political content:
- State elections: At least 19 AI-generated videos circulated during the Assam, West Bengal, Kerala, Tamil Nadu and Puducherry elections between January and April. None carried an AI-generated label. Official BJP, Congress and Dravida Munnetra Kazhagam (DMK) handles shared several of them.
- Tamilaga Vettri Kazhagam (TVK): An AI-generated video recreated the voice of former Tamil Nadu CM CN Annadurai endorsing actor and TVK founder Vijay. It went viral in March before anyone flagged it as synthetic.
- Gaurav Gogoi: Unlabelled communal deepfakes of the Congress leader, some showing him wearing a skullcap, circulated in Assam.
- Shashi Tharoor: Videos cloned the Congress MP’s face and voice to show him praising Pakistan. They remained online without labels until the Delhi HC ordered X to take them down on May 10.
Five months later, the pattern continues:
- BJP: AAP’s November 3 post quotes a BJP Delhi post that is no longer available. In January 2025, BJP Delhi posted an AI-generated video about former Delhi CM Arvind Kejriwal titled “Shark Tank: Kejriwal Version”. That post is also no longer available on X.
- Samajwadi Party: On August 31, party chief Akhilesh Yadav played an AI-generated video showing Uttar Pradesh CM Yogi Adityanath smoking a chillum at a press conference. He said the party had not created the video, but its media cell had shared it. In April, Lucknow police registered a case against party leader IP Singh over an AI-generated video depicting Adityanath as “Rehman Dacoit”.
What happened in the Chadha case?
The case Justice Jain relied on also involved AI-generated content targeting a politician who had switched parties. In July, the Delhi HC refused to order a blanket takedown of deepfakes that circulated after Chadha left AAP for the BJP. It ordered the removal of content from six flagged documents and directed X to share details of the accounts behind them. The court said his complaint appeared to concern defamation rather than a violation of his personality rights.
Gupta’s lawyers also cited personality rights cases involving Tharoor and businessman Sanjiv Goenka. The court will hear the case next on December 4.
Also Read: