On February 10 last, the Ministry of Electronics and Information Technology (MeitY) notified amendments to the IT Intermediary Rules, formally bringing AI-generated content under regulatory oversight. Effective February 20, these rules mandate clear labelling of synthetically generated information, embed permanent metadata for traceability, and compress takedown timelines to as little as three hours. The intent is laudable — protecting citizens from deepfakes, election manipulation, and non-consensual intimate imagery. The challenge lies in the execution.

India has witnessed the dark side of unregulated AI content. Deepfake videos have targeted celebrities and ordinary citizens alike, fabricated political speeches have threatened electoral integrity, and synthetic child abuse material has proliferated unchecked. By explicitly defining “synthetically generated information” as content artificially created or modified to appear authentic, the government addresses real harms. The rules cover everything from AI-manipulated videos to algorithmically altered audio, bringing platforms that enable such content squarely within India’s compliance framework.
The architecture is comprehensive. Platforms must prominently label AI content so users can immediately identify its synthetic nature. Where technically feasible, they must embed permanent metadata with unique identifiers to trace content back to its source. Crucially, these labels cannot be removed or suppressed. Significant social media intermediaries face heightened obligations — they must require user declarations before content is published and deploy automated tools to verify those declarations. For prohibited content involving child abuse, non-consensual imagery, or deceptive impersonation, platforms must act swiftly with account suspensions, content removal, and mandatory reporting to law enforcement.

These objectives have merits. Users have a right to know when content is synthetic. Victims need rapid remedies. Democracy requires distinguishing authentic political discourse from fabricated statements. The government has also shown pragmatism by excluding routine edits like colour correction and noise reduction from the synthetic content definition, responding to industry concerns. The removal of an earlier proposal requiring AI labels to occupy 10 percent of screen space reflects welcome flexibility.
LInk: https://indianexpress.com/article/opinion/columns/new-rules-to-regulate-ai-generated-content-in-india-10528053/