India mandates social media to reduce content quickly

India has ordered social media to step up policing of content and other artificial intelligence generated by AI, drastically reducing the time it takes to comply with takedown laws. It’s a move that could also change the way global technology companies process content in the world’s largest and fastest-growing Internet services market.

Changes, printed (PDF) Tuesday as updates Indian IT Regulations 2021bringing deepfakes under the regulatory framework, ordering the recording and tracking of audio and video, and reducing the platform’s compliance time, including a three-hour deadline for official removal and a two-hour window for other urgent user complaints.

India’s importance as a digital market magnifies the impact of these new laws. With more than a billion internet users and a large youth population, South Asia is a key market for platforms like Meta and YouTube, making it possible that India’s strategy can influence content management globally.

Under the revised rules, social media sites that allow users to download or share videos must disclose whether the content is artificially created, use tools to verify the claims, and ensure that deepfakes are clearly labeled and accompanied by identifiable data.

Certain categories of material – including pornographic videos, inappropriate intimate photos, and material related to serious crimes – are prohibited by law. Non-compliance, especially in cases filed by government authorities or users, can expose companies to serious liability by jeopardizing their port security under Indian law.

Laws rely heavily on systems designed to achieve this. Platforms are expected to use technical tools to verify user disclosures, identify, and deep document, and prevent the creation or sharing of prohibited content in the first place.

“The reformed IT laws represent a better way to manage AI-generated infrastructure,” said Rohit Kumar, co-founder of New Delhi-based policy think tank Quantum Hub. “Regular reporting times – such as two- to three-hour unloading windows – will greatly improve compliance and inspection efficiency, especially since non-compliance is associated with a loss of port security.”

Aprajita Rana, a partner at AZB & Partners, a leading Indian law firm, said the rules now focus on AI-recorded audio instead of all online data, while making exceptions for routine, cosmetic, or AI-related services. However, he warned that the need for mediators to remove content within three hours of knowing that it deviates from the standard conditions of free speech.

“The law, however, continues to force intermediaries to remove the information they know or receive real information, also within three hours,” said Rana, adding that the registration requirements can be used in all forms to stop the spread of child abuse and fraud.

New Delhi-based digital advocacy group Internet Freedom Foundation he said these rules are at risk of increasing censorship by over-enforcing takedown times, leaving room for people to review and push platforms to take down altogether. In a statement posted on X, the group also expressed concern about the growth of prohibited groups and the fact that it allows platforms to disclose user information to private complainants without court supervision.

“Short periods of non-compliance eliminate any personal scrutiny,” the group said, warning that the changes could weaken free speech and due process protections.

Two industry sources told TechCrunch that the changes followed a low-profile approach, with only a handful of proposals reflected in the final regulations. While the Indian government appears to have taken the initiative to limit the amount of information shared – focusing on AI-generated videos instead of all online content – other initiatives have not followed suit. The breadth of changes between the draft and the final rules required further discussion to give companies clearer guidance on what they are expected to comply with, the sources said.

The power to remove the government has long been a a controversial point in India. Social media and community groups have long been critical the scope and obscurity of the removal lawsand Elon Musk’s X challenged New Delhi in court it’s over instructions to block or remove recordsthey argue that they committed fraud and lacked sufficient immunity.

Meta, Google, Snap, X, and the Indian IT service did not respond to a request for comment.

The latest changes come months after the Indian government, in October 2025, reduced the number of authorities authorized to order content to be removed from the Internet in response to a against X’s laws on the growth and exposure of the removal forces.

The revised rules will come into effect on February 20, giving the platform less time to adjust its tracking systems. This release is in line with India’s performance AI Impact Conference in New Delhi from 16 to 20 February, which is expected to pull technology leaders around the world and policy makers in the country.