Disclaimer drama: Influencers thrown for a loop over AI labelling guidelines; compliance far off

Elon Musk prediction on AI

Influencers are scrambling to decode India’s new AI-labelling rules, and many worry that compliance could end up compromising creativity.

Over a month after the rules were introduced, Moneycontrol spoke to creators, influencers and executives of influencer marketing agencies to understand whether the new rule is being adopted. Many said the compliance requirement is turning into the one challenge no filter can fix.

The Ministry of Electronics and Information Technology, on October 22, 2025, proposed mandatory disclosure and labelling of artificial intelligence (AI)-generated “synthetic” content on social media platforms.

Since then, the mandate hasn’t trickled down in any real way, said finance content creator and entrepreneur, Shreya Jaiswal. "Barely anyone in the creator space is even talking about this."

Those who are talking about it have listed down many challenges in complying with the disclaimer.

Such a label disrupts existing design workflows and forces us to re-edit templates, transitions, and layouts, said Aditya Rana, Co- founder, 42 studios India.

"For agencies like ours, the compliance process means re-building presets, generating two versions of assets (with and without AI elements), and training creator teams. That directly increases operational cost. For solo creators, it means more manual effort and lower output volume," he said.

Globally, Ritesh Ujjwal, Co-founder, Kofluence said that YouTube introduced AI-disclosure tools but didn’t mandate a specific screen percentage, and the EU’s AI Act encourages transparent labelling but leaves the size and placement flexible.

Shivam Budhiraja, personal finance content creator and co-founder of Team Car Delight, is worried about the 10 percent AI label affecting user experience. "Such a large on-screen label, or an audio disclaimer for part of the clip, can feel intrusive and distract people from the content itself," he said.

A label might also lead viewers to misinterpret the content as false or unreliable, he added.

According to him, the rule is meant for synthetic or high-risk content, but its current wording is broad enough that everyday edits like filters, colour fixes or audio cleanup might appear to fall under it.

Another challenge is that the system relies heavily on metadata when creators do not manually declare AI involvement, Budhiraja said.

"Metadata often gets stripped when a file is cropped, compressed, screenshotted or re-edited, which makes automated detection less dependable. Even though Meta and YouTube offer options to mark content as AI-altered, barely any creators use them, so most AI-assisted content remains undisclosed," he added.

Jaiswal also pointed to platform bias. She said that platforms like Meta openly discourage any kind of persistent label or logo in Reels. "Adam Mosseri had said that Meta deprioritizes branded or overlaid content in the algorithm. So, the moment you add a 10 percent AI label, it’s highly likely your reach takes a hit, because the platform doesn’t see it as organic creator-first content anymore."

She also said that a lot of Reels are just 8-10 seconds long and an AI label covering 10 percent of the screen for a fraction of that time will affect the framing, storytelling and the whole flow.

While the intent is positive, a rigid 10 percent rule may feel limiting for certain creators, especially in visually rich or cinematic content, Ujjwal added.

Brands that were excited about AI-powered campaigns are now slowing down to understand what qualifies as AI content and what does not, said Praanesh Bhuvaneswar, Co-founder and CEO, Qoruz.

One example is multilingual campaigns. "Over the last six months, brands were heavily using AI to create language versions of the same video. They would shoot one main reel and then use AI voice cloning or lip-sync adjustments to generate four or five regional versions. It saved both time and money," Bhuvaneswar said.

But he noted that with the new rule, every regional version now technically needs an AI label, which changes the visual aesthetic. "Some brands are choosing to reduce the number of language versions for this reason, impacting production decisions"

Another instance is AI product demos. Bhuvaneswar pointed to technology, D2C (Direct-to-consumer), and mobile brands using AI to build clean product explainer animations that look polished. "These were not deepfakes. But now, teams have to rethink how to label such content without distracting from the information."

He recently saw a brand request a re-edit of all their reels because the AI label covered their Call to Action (CTA are prompts on the screens like sign up or buy now ) in the first two seconds. AI labels sometimes sit on top of these cues, which forces creators and brands to re-edit because the message gets blocked.

Bhuvaneswar added that agencies are now building internal checklists where every Reel must go through an “AI touchpoint review” to check whether voice, visuals, backgrounds, or transitions used generative tools.

"We don't have reliable tools to automatically detect what's AI-generated and what's not," said Arsh Goyal, AI and technology creator, adding that the technology is not very mature yet.

42 studios India's Rana added that most of the work to add AI labels is manual. "AI tools don’t yet have a uniform way to auto embed compliant labels. This slows down production pipelines especially for creators who mass-produce reels, shorts, or ad creatives."

• Based on their production cycles, Rana said that adding the required AI label is increasing editing time by 18 percent–32 percent
• Revision cycles by 1.4 times
• Post-production cost by 12 percent–20 percent depending on the complexity of the content. For creators publishing 20–40 videos/month, adding AI labels translates to an additional 6–9 hours of monthly editing, Rana said.
• For agencies managing bulk ads, it increases project timelines by 1–2 days per campaign, he added.

To implement AI labels, agencies must create additional supers and overlays for multiple aspect ratios, 9:16, 16:9, 1:1, and more, depending on where the content is deployed. This means extra design and production hours, said Aayush Bansal, Co-founder, Black Cab, an integrated & Digital Marketing Agency.

Larger agencies and enterprises also need a new legal-review cycle to ensure every asset is compliant, he added.

For creator Amtul Faheem who makes videos on everyday recipes and home-style cooking, the AI-labelling mandate makes him keep things real.

However, when it comes to adoption of the new rule across the creator community it is still low and scattered, Rana said.

Large brands and organised media houses have begun transitioning, but the creator economy, especially independent creators, is still far from consistent compliance, he added. "Many are still unaware of the exact guidelines, and others are waiting for platforms like Instagram and YouTube to provide built-in compliance tools."

"It's not being widely adopted because the rules aren't final yet," Goyal said.

Content creator Chandan N said that a standard set of guidelines will help all creators apply it consistently.

Jaiswal added that compliance is low and will likely remain low because of a lack of awareness within the creator community. There are also several unanswered questions, she said.

"What qualifies as AI-generated content?" she asked. "If my reel is entirely shot and edited by me but I just used auto-captions or suggestions, does that still count? If I use Canva’s smart text generator for two words in a carousel, is that AI content?"

"What’s the format of the label? Should it be verbal, visual, or both? Does it appear once, or is it persistent throughout the video? Who is monitoring the labeling of AI content and what’s the penalty for non-compliance? "In a country like India, unless there’s a fine or an actual enforcement mechanism, no one is going to follow it," Jaiswal said.

Rana recalled a recent internal project where his team audited 112 AI-assisted videos. "Over 65 percent needed rework because the current guidelines are interpreted differently by every editor. There’s confusion on whether the label must run for the full duration, how to handle partially AI-generated clips, and what standard design/placement to follow," he said.