The IT Ministry has issued revised guidelines requiring social media platforms such as Facebook, Instagram, and YouTube to label all AI-generated content and take down deepfakes within three hours of receiving a government or court order. The directive, announced Tuesday, responds to the spread of synthetic media online and applies to any platform operating in India.
The rules require platforms to ensure AI-generated material carries embedded identifiers or metadata. Platforms are also barred from allowing users to remove or suppress those labels once applied.
According to the MeitY order, social media companies must deploy automated tools to detect and prevent the circulation of illegal, sexually exploitative, or deceptive AI-generated content. The guidance references the Bharatiya Nyaya Sanhita, 2023; the Protection of Children from Sexual Offences Act, 2012; and the Explosive Substances Act, 1908 as legal baselines for what counts as a violation.
What platforms must do
The notification asks intermediaries to act "expeditiously" when they become aware of any rule violation related to synthetic content. That includes content that is created, generated, modified, hosted, displayed, uploaded, published, transmitted, stored, updated, or shared on their services.
Under the updated guidelines, platforms must also use "reasonable and appropriate technical measures, including automated tools or other suitable mechanisms" to prevent users from disseminating synthetic content that violates Indian law. The order does not specify which tools should be used, leaving the technical approach to the platforms.
Users will see a new obligation as well: the draft rules mandate disclosure when posting AI-generated or modified content. Platforms must adopt technology to verify those declarations. Several social media companies already offer features that allow users to label content as generated by AI.
Ongoing notice to users
The order also requires platforms to inform users about the consequences of AI misuse at least once every three months, through their rules, privacy policies, user agreements, or other appropriate channels.
An intermediary shall periodically inform its users, at least once every three months, in a simple and effective manner through its rules and regulations, privacy policy, user agreement, or any other appropriate means, the notification said.
Once a platform receives a formal order or court direction flagging deepfake content, the three-hour removal clock starts.
Why this matters for government employees
Government staff who administer public information, approve content, or rely on official statements face direct consequences from this rule. If a synthetic version of a minister's speech, a departmental notice, or a public safety announcement circulates online, the new rules give you a clear window to have it removed when the government flags it.
Officials who oversee social media accounts for departments need to know their own posting practices are now covered by the same labeling requirements. An original photo posted by an official account might count as synthetic content if edited with AI tools, and the account could field legal consequences for failing to label it.
For those using AI to draft reports, press releases, or visual materials, the standard is now legal rather than editorial. Failing to label synthetic content is not just a caption choice; it is a direct violation of a government order with a fixed enforcement timeline.
Government employees looking to understand the broader national policy on AI may find value in the AI for Government training resources, which track how public-sector institutions are adopting the technology.
Your membership also unlocks:
