The Ministry of Electronics and IT (MeitY) is working on an artificial intelligence (AI)-specific law, under which it may not prescribe any penal consequences for violations, in a recognition that the technology’s benefits outweigh its downsides, The Indian Express has learnt. The law, which could be released as a standalone piece of legislation, could require social media platforms like Facebook, Instagram, YouTube, and X to add watermarks and labels to content that has been generated using AI. The ministry is also looking at the legal parameters under which it can ask companies building large language models to train their databases on Indian languages and “India-specific” content so that the models can understand the “Indian context” better, a senior government official said. This is being looked at after some AI systems, like Google’s Gemini, were found to be offering different answers for the same kind of queries involving world leaders and Prime Minister Narendra Modi last year. The IT Ministry did not respond to a request for comment. Last November, Union IT Minister Ashwini Vaishnaw had said that the government will form regulations to control the spread of deepfakes on social media platforms, terming them a “new threat to democracy.” The election cycle earlier this year stopped work on the law, which is now understood to have found newfound momentum within the ministry. At the time, Vaishnaw had said the Centre’s plan will have four key pillars: detection of deepfakes, their prevention by removing or reducing their virality, strengthening reporting mechanisms, and spreading awareness about the technology. In May, IT Secretary S Krishnan during an industry event had said that while the government will attempt to regulate AI, it will not do so at the cost of innovation, signalling the fact that the ministry is expected to take a middle path for AI regulation, as it has taken in some of its recent technology laws such as data protection, where the focus is not just on protecting civil rights but also on facilitating the country’s growing start-up base. “While we will attempt to regulate AI, we are clear that innovation is not stifled in the process. It (innovation) needs to be encouraged… Like we did with the DPDP (Digital Personal Data Protection) Act, we will ensure that both the interests of innovation and protection of vital interests will come in in the future,” Krishnan added. To be sure, some social media and AI companies have said that they have started watermarking AI-generated content on their platforms before regulations require them to do so. Adobe, Arm, Intel, Microsoft, and Truepic have also created a joint project called the Coalition for Content Provenance and Authenticity (C2PA), which focuses on systems to provide context and history for digital media. Meta has said that it is building tools that can identify “invisible markers” in content pieces relying on standards set by the C2PA so it can label images from Google, OpenAI, Microsoft, Adobe, Midjourney, and Shutterstock.
