Used AI to clean up audio, add subtitles, or prepare a presentation? You may not have to label it as ‘synthetic content’ after all. A day after tightening rules for social media platforms, the government has clarified that AI used in good faith, especially for education, training, or technical improvements, will not automatically be treated as synthetic-generated content (SGI). The clarification came through a set of FAQs issued by the Ministry of Electronics and IT (MeitY), following amendments to the IT Rules, 2021. The new rules will take effect from February 20, 2026. What actually counts as synthetic content? The government says not every AI-assisted edit or creation falls under SGI. In its clarification, MeitY said: Not every AI-assisted creation or editing qualifies as SGI. Content is treated as SGI only when it is artificially or algorithmically created or altered in a way that it appears real or authentic or true and is likely to be indistinguishable from a real person or real-world event.
In simple terms, content must be labelled as synthetic only if it is designed in a way that could mislead people into believing something fake is real. So if AI is being used to generate a deepfake video that looks like a real person said or did something they never did, that clearly falls under SGI. But routine editing does not.
What has changed?
Under these updated rules: These changes come after growing concerns about deepfakes, fake videos, impersonation, and AI-generated misinformation spreading online.
Also read: Govt orders social media platforms to remove deepfakes within 3 hours
What AI use is exempt? The FAQs clearly state that routine or good-faith use of AI will not require synthetic labelling, as long as the original meaning is not changed.
Examples that are exempt: The FAQ further states: The use of computer resources solely for improving accessibility, clarity, quality, translation, description, searchability, or discoverability shall not be treated as SGI, provided the process does not generate or alter or manipulate any material part of the underlying content. Also read: Russia attempts to fully block WhatsApp, says messaging platform
When will AI content be considered unlawful? The government has drawn a clear line. If AI is used to create fake certificates, forged IDs, fake official letters, fabricated electronic records, and deepfake videos impersonating real people, then such content will not be exempt. It may be treated as unlawful SGI or false records.
In short, the message is clear: using AI to improve clarity or accessibility is fine. Using it to deceive is not.
