India’s New AI Rules Explained: What Social Media Platforms Must Do Now

The Indian government has announced a new set of rules to tackle the misuse of artificial intelligence (AI), especially deepfakes. The changes, introduced through amendments to the Information Technology (IT) Rules, 2021, place stricter responsibilities on social media platforms and aim to protect users from fake AI-generated content.

Here are the biggest changes under the new rules:

1. AI-generated content must be clearly labelled

Social media platforms and intermediaries must ensure that AI-generated content carries clear labels and traceable metadata. This will help users identify whether a photo, video or audio clip has been created using artificial intelligence.

2. Illegal AI content must be removed within 3 hours

One of the biggest changes is the shorter deadline for taking down unlawful AI-generated content. Platforms must now remove such content within three hours after receiving a valid order from the government or a court. Earlier, they had 36 hours.

3. Faster action on user complaints

The government has also reduced the time platforms get to handle complaints.

• Sensitive complaints such as AI-generated nudity, impersonation and non-consensual intimate images must be addressed within 2 hours.
• Other grievances must be resolved within 36 hours, instead of the earlier 72 hours.

4. Platforms must stop harmful AI content

Social media companies must deploy automated tools or other technical systems to detect and prevent users from creating or sharing AI-generated content that violates Indian laws.

The rules specifically cover deepfakes, child sexual abuse material, impersonation, non-consensual intimate imagery and other harmful AI-generated content.

5. Stronger accountability for social media companies

Platforms must also make users aware of the legal consequences of posting unlawful AI-generated content. They are expected to strengthen their moderation systems and improve compliance with Indian laws.

6. Failure to comply can invite legal action

If platforms fail to follow these rules, they may lose the legal protection available under Section 79 of the Information Technology Act. This means they can face legal action for unlawful content hosted on their platforms.

Government also investing in deepfake detection

Along with the new rules, the government said it has approved 13 Responsible AI projects under the IndiaAI Mission to develop indigenous deepfake detection tools. These include projects by IIT Jodhpur, IIT Madras and IIT Kharagpur.

The Centre also said cyber security awareness efforts have expanded significantly, with 6,650 awareness workshops conducted across the country, benefiting more than 11.37 lakh people.

The information was shared by Minister of State for Electronics and Information Technology Jitin Prasada in a written reply in the Lok Sabha.

Read More: Pakistani Handle’s Deepfake Video Of Delhi Cop Cites Amit Shah; Account Withheld