Institute of Legal and Management Studies

The Supreme Court today refused to entertain a public interest litigation assailing authorities' failure to tackle Artificial Intelligence (AI)-generated deepfakes and seeking constitution of a Court-monitored Expert Committee to draft model AI regulation law.

A bench of Justices Surya Kant and N Kotiswar Singh heard the matter and relegated the petitioner, a lawyer, to the Delhi High Court, which is dealing with the issues and has been passing orders from time to time.

"We do not deem it necessary to entertain these parallel proceedings. Petitioner is relegated to Delhi High Court with liberty to seek his impleadment as intervenor and assist. We request the High Court to accord audience to the petitioner and consider the valuable suggestions as may be given by him", the Court ordered.

Notably, during the hearing, the petitioner (in-person) expressed concerns about a deepfake video of Colonel Sofiya Qureshi, who became the face of India's 'Operation Sindoor' after she gave press briefings on the same. He also highlighted that the government had assured to come out with some laws regarding deepfakes.

In response, Justice Kant remarked, "You are a member of Bar Council, right? Your problem appears to be, you want to go outside, some media people are standing there and you want to speak before the mic".

Subsequently however, the bench impressed upon the petitioner that it was not saying that the issue ought to be rejected outrightly. "What we are saying is, Delhi High Court is already examining it from last couple of years. Join there. So many steps have been taken by High Court...If Union's stance or High Court directions are not satisfactory, you can always come here", Justice Kant said to him.

The judge also opined that there is no point in debating the issue on public platforms. "These cyber criminals are so smart, they will create another such video by the time you step out of the court. Something will have to be done very seriously. Go and assist High Court, maybe you have some new ideas."

The petition, filed by Advocate Narendra Kumar Goswami, sought directions for the Ministry of Electronics & Information Technology to frame Rules under IT Act, 2000 mandating:

(i) Watermarking of all AI-generated content (images, audio, video) with metadata disclosing origin, tools used, and creator identity, as per China's Deep Synthesis Provisions;

(ii) A 24-hour takedown mechanism for deepfakes, mirroring Rule 3(1)(b)(vii) of IT Rules, 2021 (CSAM26 protocols), with penalties under Section 45 of the IT Act for non-compliance;

(iii) Algorithmic audits of AI platforms quarterly, conducted by CERT-In empanelled auditors (CIAD- 2024-0060.

It further sought directions for establishment of an AI Regulation Body under Section 88 of the IT Act, chaired by a retired Supreme Court judge, with members from NITI Aayog, CERT-In, and academia, to oversee compliance.

As against the Election Commission of India, the petitioner sought direction for establishment of a Deepfake Monitoring Cell under Article 324 with powers to (i) Pre-certify all political advertisements using AI tools; (ii) Issue real-time take down orders to platforms under Rule 16 of Conduct of Elections Rules, 1961; and (iii) Maintain a public repository of debunked deepfakes.

It was further prayed that the Model Code of Conduct be amended to prohibit undisclosed AI-27 generated content during elections, enforceable under Article 324, with penalties under Section 171G IPC for violations.

Qua the Ministry of Home Affairs, the petitioner sought the following reliefs:

- Direction to develop within 90 days a National Protocol on AI threats under Section 66F IT Act (cyber terrorism), including (i) A dedicated cyber-forensics unit under NIA to investigate foreign-origin deepfakes; and (ii) Mandatory reporting of deepfake incidents by platforms to CERT-In under Section 70B(5) IT Act.

- Implementation of training modules for police on deepfake detection and FIR registration under Sections 419/500 IPC, in collaboration with NICFS.

The petitioner also prays that the Ministry of Education launch a National Deepfake Literacy Mission under Samagra Shiksha Abhiyan, integrating AI/digital literacy in NCERT curriculum (Class VI-XII) within 12 months, with funding under Article 21A. Apart from this, he seeks the following declarations:

- That the respondent-authorities' failure to regulate AI-generated deepfakes violates (i) right to privacy/dignity under Article 21 (ii) Voters' right to truth under Article 19(1)(a) and (iii) Right to equality under Article 14;

- That the IT Act is inadequate to address deepfakes, necessitating judicial guidelines under Vishaka v. State of Rajasthan (1997).

In addition, he demanded constitution of a 5-member Court-monitored committee chaired by a retired SC Judge, with members from DG-CERT-In, ECI Secretary, DG-NIA, and Director-IIT Delhi to draft model AI regulation legislation.

Case Title: NARENDRA KUMAR GOSWAMI Versus UNION OF INDIA AND ORS., W.P.(C) No. 300/2025