India’s New AI Labelling Rules: Internet Game-Changer
The digital era is about to shift. In a bold move, the Ministry of Electronics & Information Technology (MeitY) in India has proposed sweeping amendments that require clear labelling of AI-generated content, redefining how the internet works in the world’s largest democracy. These new rules aren’t just tweaks—they herald a major transformation in the relationship between creators, platforms, regulators and users.
In this blog we’ll explore:
- What exactly the rules are
- Why they’ve emerged now
- Who will be affected and how
- What the broader implications are for the internet, for creators, for users, and for digital business
- What critics and supporters are saying
- How you (as a user, creator or business) should prepare
- FAQs to clear up the major questions
By the end, you’ll understand why this change isn’t just “another regulation”, but a real potential inflection point for digital content in India — and globally.
1. What Are the New Rules?
In October 2025 the Indian government proposed draft amendments to the existing digital media / intermediary rules. These revolve around what is called “synthetically generated information” (i.e., content made or modified using AI) and the obligations on platforms, creators and users to label such content.
Key requirements include:
- Platforms must ask users when uploading whether the content is AI-generated or AI-altered.
- If content is AI-generated or altered, a visible label must be applied. For visual media, the label must cover at least 10% of the surface area of the image/video; for audio, it must appear in the first 10 % of the clip.
- Platforms must deploy “reasonable and appropriate technical measures” (for example automated tools) to verify declarations of AI content.
- Platforms may lose their legal “safe harbour” (i.e., intermediary immunity) if they fail to comply with these labelling obligations and verification duties.
- Definition: “Synthetically generated information” is defined as information that is artificially or algorithmically created, generated, modified or altered using a computer resource, in a manner that appears reasonably authentic or true.
- The draft rules are open for public and industry feedback (with a deadline of 6 November 2025).
Why these specific numbers (10%, etc.)?
The requirement for the label to cover at least 10% of the visual surface or first 10% of duration is unusual and specific — the government is introducing quantifiable visibility standards, among the first of their kind globally.
In short: if you upload a video that’s fully AI generated (or significantly altered via AI), you’ll have to clearly mark it, visibly and audibly, so users can tell it’s synthetic.
2. Why Are These Rules Emerging Now?
Several factors converge to push this regulatory shift.
Escalation of AI-generated content and deepfakes
The rise of generative AI tools means it’s increasingly easy to create images, videos and audio that look authentic but are not. These “deepfakes” pose risks: impersonation, misinformation, reputational damage. In India, a viral deepfake video of a Bollywood actor put these concerns in the spotlight.
Domestic socio-political risk
With nearly a billion internet users, diverse languages, religions and regions, India faces a heightened risk of AI content being weaponised for social unrest or communal mischief. The government explicitly referenced the “growing potential for misuse” of synthetic media.
Global regulatory momentum
India isn’t alone. Other jurisdictions — notably the AI Act in Europe and China’s wiring of labelling rules — serve as models and pressure. India wants to align with global standards and show leadership.
Technology and platform responsibilities
Platforms like Meta, Google and others already have labelling practices but India’s proposals push for proactive verification, not just reactive labelling when something is flagged.
3. Who Will Be Affected and How?
Platforms & intermediaries
Major social media, video and content platforms will have new compliance burdens:
- Asking for declarations from users
- Deploying technical tools to detect synthetic content
- Clearly labelling AI-generated/altered media
- Potentially losing safe harbour protection if they don’t comply
This shift means compliance costs will rise, operations will need changes, and risk exposure increases.
Creators & publishers
If you’re a creator, influencer or publisher producing AI-generated or AI-altered content, you’ll need to clearly mark that content as synthetic. Failure to do so may lead to removal, takedowns or other liabilities.
Users / general public
For everyday users, the rules mean:
- More transparency: you’ll know which content is synthetic
- Potentially fewer wild deepfakes unchecked
- Platforms may start flagging or labelling AI content, which may alter how you consume content
Innovators & AI tool providers
Companies building generative AI tools or providing them in India will face pressure to build in “label at source” features, watermarking, or metadata embedding to ensure compliance upstream.
Businesses and advertisers
For brands and advertisers who use AI-generated assets (e.g., deepfake-style marketing, synthetic influencers), the rules mean you must ensure labelling and transparency. Non-compliance may hit brand reputation or legal risk.
4. Broader Implications: The Internet Game Has Changed
Let’s explore the ripple effects: some expected, some less obvious.
Transparency becomes baseline expectation
The default assumption might shift from “everything is real or unchanged” to “some content is synthetic – check the label”. As users become used to labelled synthetic content, trust paradigms shift.
Increased friction in content creation
Label-and-verify adds friction. Creators will need to manage labelling, maybe add visible markers. Platforms will need technology to verify. This may slow some types of viral content, or alter how quickly things are published.
Incentives shift for AI use
If synthetic content must be labelled, some creators might avoid making it (or use less obvious synthetic tools). Others will lean into transparency as a badge of trust — “Yes, this was AI-generated”. The public may reward honesty.
Emergence of new “watermarking/metadata” tech
The rules explicitly push platforms to embed metadata or unique identifiers in synthetic media. That opens the door for watermarking tech, provenance tracking, forensic detection services. India may become a market for such tech.
Impact on misinformation & deepfake abuse
By making synthetic media visible and traceable, authorities hope to reduce malicious deepfakes. That may mean fewer identity-theft or impersonation incidents, but it also means new tactics may emerge to bypass labelling.
Global precedent and export effects
India’s rules may become a blueprint for other jurisdictions, especially in the Global South. Platforms operating globally will need to adapt — and India may attract AI governance talent or regulatory tech firms.
Content moderation and platform liability increases
Platforms may have to invest more heavily in moderation, detection, and compliance. Smaller platforms may struggle or get acquired. The dynamic of “self-regulation” shifts toward stronger state oversight.
Creative and marketing impact
AI-generated art, influencers, synthetic voices, deepfake-style marketing will need clear labelling. Brands may pivot to “Certified human vs synthetic”. The boundary between real and synthetic becomes a marketing tool.
Legal and free-speech tensions
While the intention is to curb misuse, there are questions about what happens to satire, parody, artistic uses of AI, and freedom of expression. Labelling mandates may impact how creative media is distributed.
5. What Supporters and Critics Are Saying
Supporters argue:
- These rules bring accountability and transparency to a landscape of rampant deepfakes.
- They protect individuals, elections, reputations from misuse of synthetic media.
- They align India with international best practice and show regulatory maturity.
Critics warn:
- The labelling requirement may stifle innovation or impose heavy compliance on smaller platforms and creators.
- Enforcement may be inconsistent, or tech burdens may favour large corporates.
- There is potential for overreach: what counts as “synthetic”? Will human-edited content get caught? Could parody be unduly burdened? Some experts cautioned that balancing authenticity and accountability with freedom of speech “will be key to the success of this framework”.
- Technical verification is not trivial — false positives/negatives may appear, raising risk of censorship or abuse.
6. What Should You Do? (Users, Creators, Businesses)
As a content creator / publisher:
- Start auditing: Do you use AI-generated or AI-altered content?
- Plan for labelling: ensure any synthetic piece is clearly marked and the label covers visible/audible criteria.
- Stay updated: the draft may evolve when finalised; keep track of notifications from MeitY.
- Build provenance: embed metadata, keep logs of how content was generated, to show transparency.
As a platform/operator:
- Update terms of service: clearly ask users about synthetic content upload.
- Build/outsource detection tools for synthetic media verification.
- Ensure UI/UX supports visible labelling (10% surface/first 10% duration).
- Train moderation teams to handle synthetic-content rules and appeals.
- Engage with policy: provide evidence in feedback to the draft, shape how rules apply to you.
As a user/consumer:
- Become a bit more skeptical: check for labels “AI-generated” when watching videos or seeing strange content.
- Appreciate transparency: look for labelled synthetic media instead of assuming everything is “real”.
- Keep informed: major platforms will update their policy pages and transparency reports.
As a brand/advertiser:
- If you use synthetic influencers, voices or visuals — ensure compliant labelling and transparency.
- In your marketing strategy, highlight “human vs synthetic” where relevant — brands may gain trust by clarifying authenticity.
- Monitor risk: the cost of non-compliance or reputational backlash can be significant.
7. FAQs (Frequently Asked Questions)
Q1: What qualifies as “synthetically generated information”?
A: Under the draft, it is “information that is artificially or algorithmically created, generated, modified or altered using a computer resource, in a manner that appears reasonably authentic or true.”
In simpler terms: if something is made or significantly changed with AI in a way that could mislead someone into thinking it’s fully authentic, it needs labelling.
Q2: Does this apply to text-based AI (chatbots, AI writing) or only images/videos?
A: The draft explicitly covers synthetic media (images, audio, video) but also defines “information” broadly — so text generated or modified by AI could fall under scrutiny. The biggest visible obligations (10% label size) are for visual/audio. Text labelling may require appropriate notices.
Q3: When will these rules come into force?
A: They are currently in draft form and opened for public and industry feedback until 6 November 2025. The final notification date is yet to be announced.
Q4: What happens if a platform fails to label synthetic content?
A: The platform may lose its safe-harbour protections under the existing IT Rules (Intermediary Guidelines). This means it may be held responsible for third-party content.
Q5: Will this kill creative uses of AI for art and entertainment?
A: Not necessarily—but it will require transparency. Creative works that use AI will need to be labelled so users know the origin. The challenge will be in how strictly enforcement is applied and how well creative/transformative uses are distinguished. Critics warn about chilling effects.
Q6: How will the “10% label” for visuals/audio work in practice?
A: For an image/video, the visible label must cover at least 10% of the surface area. For audio (or video audio), the label must appear during the first 10% of its duration (e.g., the first 10 seconds of a 100-second clip). The label or identifier must be permanent and cannot be removed by intermediaries.
Q7: Will non-Indian platforms (global social media) also have to comply?
A: Yes, if they operate in India as an “intermediary” or significant social media intermediary they will be subject to the draft rules and associated obligations. Platforms servicing Indian users will likely need to adapt.
Q8: What about satire, parody or fictional AI content?
A: The draft does not exempt satire explicitly. The challenge will be in how regulators interpret “reasonably authentic or true”. Creators of parody or fictional AI may need to ensure clear disclaimers to avoid being incorrectly labelled or flagged.
Q9: Does this mean the end of deepfakes?
A: Not quite. The labelling regime aims to reduce misuse by making synthetic media transparent, but technology will continue evolving. Enforcement, detection, user awareness and accountability will all still matter. This is a step—not a full stop.
Q10: How might businesses benefit?
A: Businesses that adopt transparency and trust-building early (e.g., marking synthetic content clearly) can gain consumer confidence, avoid regulatory risk, and position themselves as compliant digital-first innovators. Moreover, enterprise tool providers for watermarking, content provenance & detection may find growth opportunities.
8. Conclusion: A New Era for the Internet
With the proposed rules from India’s MeitY, the internet is poised for a structural change. The era when anything could be uploaded and assumed “real” is shifting toward one where synthetic content is visible, auditable and accountable.
For creators, platforms and users alike, the message is clear: the future of content is hybrid (human + AI) — but if it’s synthetic, be upfront about it. Transparency will be the new standard.
For India, this is a bold move to safeguard its digital ecosystem, trust in media, and protect individuals and society from AI-driven harm. If properly implemented and enforced, it could set a global standard.
For you — whether you’re creating, consuming, marketing or regulating — this is your moment to adapt. The internet as we know it is changing; get ahead of the curve.