AI? Say so

Growing up, we had a ritual at home. In those days, VCPs (video cassette players) were rented (as owning one seemed highly unrealistic), and an entire day would be dedicated to watching VHS (Video Home System) tapes. I remember being bathed and fed and plonked in front of the TV and Imran-da (the neighbourhood movie shop owner) setting up the VCP, carefully stacking recent hits one on top of the other. The whole day was spent in wanton cinematic pleasure as our small family and sometimes neighbours partook in the movie marathon. The last movie of the day would be a horror film reserved for my father after he’d get home from work. My father enjoyed scary movies, and it’s safe to say nothing really spooked him. Since I wanted to be much like my dad, we’d sit and watch those films together. There I was, all of 6-7 years old, sitting sometimes wide-eyed, sometimes eyes shut, forcing myself to watch films such as ‘Purana Mandir’, ‘Veerana’, ‘Bees Saal Baad’, among others.

No one else watched but dad and I; no one stopped me from watching either (bad parenting if you think about it). But it ingrained in me a love for horror; a kind of daring that gives the most delightful adrenaline rush (ask other horror lovers and they’ll say the same). Along with horror films, obviously came the interest in Halloween — an entire day dedicated to costumes, thrills, and spine-chilling horror; what’s not to love?! This year, I finally put into action a long-held wish — a mini-Halloween bash at home. Ghosts, ghouls, and cheeky pumpkins decorated my home while yummy treats were generously distributed to neighbourhood kids. Dressed as a wicked devil, I was possibly the only adult in the neighbourhood as enthusiastic about Halloween as the children. My costume, put together with minimal effort, was quite the hit, and sharing it on Instagram brought on scores of compliments on my make-up, while some were genuinely frightened. But here’s the catch — my picture was enhanced by Meta AI, turning me into a scary zombie devil. Without an AI tag, though, I could bask in all the glory and take credit for something that I didn’t do. For me, this was a social experiment as I planned to write about it here in my column.

The Ministry of Electronics and Information Technology (MeitY) recently released draft amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, to control the abuse of generative AI and the proliferation of deepfakes. The amendment makes artificial intelligence (AI) tags mandatory in published content and explains “synthetically generated information”, mandating platforms, especially sizeable social media intermediaries (SSMIs) that have over 5 million registered users in India, such as Facebook, Instagram, Snapchat, and YouTube, to mark such content through metadata and visible or audible labelling. If any social media platform allows AI-generated content, then, as per rule 3(3), it must be “prominently labelled or embedded with a permanent unique metadata or identifier”.

Early use of AI was often done surreptitiously, but the shame or guilt associated with it is now long gone. Today, many businesses chest-thump their use of AI and boast of higher productivity, faster project turnarounds, and happier human resources. As AI adoption continues to skyrocket, putting down some rules may not seem that out of place. India is one of the leading adopters of AI, and while the government has no immediate plans to introduce an AI law, the India AI Governance Guidelines (which is separate from the one urging for AI labelling of social media content) now looks to encourage innovation with some balustrades. With emphasis on trust, safety, accountability, equity, and sustainability, the government’s approach has evolved from minimising AI risk to a hands-off approach toward enabling adoption.

The move to label AI content has sent ripples, especially among creative folk, who have pleaded to be kept out of the ambit. The creative industry argues that a blanket law forcing them to show labels covering 10 per cent of screen space or duration is a regulatory overreach. The industry further contends that such a disclosure would dent the use of AI in visual effects, animation, and communications. While the creative industry rallies for alternate kinds of disclosures, any sane person who has been knocked down by deep fakes would vouch for MeitY’s rules. Declaring AI use, especially where it impacts consumer trust and behaviour, should be mandatory. Recently, several leading actors such as Amitabh Bachchan, Aishwarya Rai Bachchan, Abhishek Bachchan, Anil Kapoor, Akshay Kumar, and Hrithik Roshan have sought court injunctions against the misuse of their identities online. Deep fakes have affected both celebrities like Rashmika Mandanna and the common people alike. The latter became victims of several kinds of financial fraud perpetrated using deepfake technology. A McAfee Labs report warned that almost 39 per cent of Indian consumers have encountered deepfakes during major sale events. Indians have lost a whopping Rs 22,845.73 crore to cyber fraud in 2024; a 204 per cent jump since 2023. Now back to my innocuous Halloween picture. I could mislead people into thinking that I do, in fact, have deft make-up skills; I could profit from it or scam people. Given the context of fraud prevention, the AI labelling diktat seems cautionary and necessary.

Views expressed are personal. The writer is an author and media entrepreneur