India, May 22 -- The Bajaj Group, with interests in two-wheelers, finance and electricals, recently launched an ad film commemorating its 100-year journey, inspired by the values of its founder Jamnalal Bajaj, a freedom fighter and social reformer, close to Mahatma Gandhi. The film that uses Artificial Intelligence (AI) to bring Jamnalal Bajaj and Gandhi alive on screen was created by marketing-technology company Wondrlab and directed by filmmaker Rajkumar Hirani. It combines old-world storytelling with new-age technology reflecting the Group's journey of preserving values while embracing the future. Since brands are rapidly integrating AI into campaigns, influencer marketing and personalised promotions, it's no surprise the Advertising Standards Council of India (ASCI) has released draft guidelines for responsible labelling of AI-Generated or synthetic content in advertising. Aimed at transparency and consumer protection, it aligns with the amended Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2026. The draft spells out high, medium and low risk ads. The high-risk ads infringe on rights, make misleading claims, or violate the ASCI Code and are prohibited even if an AI label is used. These include fabricating testimonials, exaggerating product results, using deepfakes, copyrighted work or a person's likeness without consent, or using AI to generate fictional authority figures such as a fake doctor to promote a product. Medium-risk ads are those where AI use materially influences consumer decisions and require labelling. Examples include virtual or synthetically generated influencers and ambassadors, replicating a real person's likeness or voice even with their consent for personalised messaging, using AI for paid or sponsored product suggestions which must specifically be labelled as 'sponsored by'. Low-risk ads require no labelling as they follow routine editing, colour correction, noise reduction, standard blemish removal or use AI in ways that have no material impact on a consumer's ability to make an informed choice. Kumar Awanish, group chief operating officer, Cheil SWA, said ASCI's categorization of high, medium and low risk AI ads seemed fine, though nitty-gritties must be fixed when the ad industry sends it feedback. Clarifications are needed on who or what decides whether AI use is "materially influencing" consumer decisions in medium-risk ads. Or, when no labelling is required what's an acceptable level of modifications, Awanish said. Ambika Sharma, founder of Pulp Strategy, too, broadly agreed with guideline architecture. "The instinct is right. A blanket label on every ad that touched AI would be useless. Consumer label fatigue is real, and once a label is used for everything it loses meaning. A risk-based approach forces the industry to think about consumer impact, not technology," she said. But a clearer demarcation is required between low-risk and medium-risk ads. Routine editing, colour correction, ambient effects, fantasy elements etc. don't need labelling. "But what about a fully AI generated background in an otherwise real shoot, or an AI assisted voice that is not a replica but synthetic? What about a real model with AI altered features? The draft is only directional, the final guidelines must be operational," Sharma said. Also, though paid AI recommendations require a "sponsored by" label, but as AI-driven product suggestions inside chatbots, voice agents and search experiences increase, the rule must hold up when the interface is conversational, not just visual, she added. AI usage in ads poses a risk and pretending otherwise would be dishonest, Sharma said. "But the risk is not AI. The risk is intent. AI does not invent a misleading claim. A brief does. AI does not decide to use a celebrity's face without consent. A team does," she added. Among specific risks, she called out deepfakes and unauthorised likeness, synthetic product demonstrations with misleading outcomes, AI generated medical or finance experts who exploit vulnerable audiences and cause real harm. "In almost every problem case AI did not create a new violation. It amplified an old one," Sharma said. Bigger questions that need addressing are around liability when AI is used across a chain of vendors. Not just that. What happens to work already in the market with brands running hundreds of AI creatives? "The draft tells you how to label new work but the transition rule is missing," Sharma argued. No serious brand or agency will publicly contest the fact that consumers deserve transparency on synthetic content. "But tech is changing at supersonic speed and regulation has to evolve accordingly," Cheil's Awanish said....