The Ministry of Electronics and Information Technology (MeitY) has proposed draft amendments to the IT Rules, 2021, to regulate synthetic content, including AI-generated videos, images, and voices. The draft aims to make the creators and platforms behind such content accountable and transparent. If adopted, India would become one of the first democracies to formally address the dangers of AI-driven misinformation. Deepfakes and Democracy.
Features of Draft Amendments
The government’s goal is to curb the spread of impersonation, fake news, and deepfake-based fraud without stifling innovation.
- They define “synthetically generated information” as content created or altered by algorithms to resemble authentic media.
- They require platforms that create or host such content to label it clearly, for example, dedicating at least 10 per cent of visual space or the first 10 per cent of audio to disclaimers.
- They mandate automated detection systems and user declarations for synthetic media uploads.
- They preserve safe-harbour protection for intermediaries that remove harmful synthetic content, while penalising those that don’t.
Challenges/Needs for Amendments
- The deepfakes have infiltrated politics, entertainment, and social discourse.
- The AI technology can also destroy reputations, manipulate elections, or incite violence.
- India’s digital population is vast, multilingual, and heavily reliant on social media for news. The risk of viral misinformation is therefore exponentially higher.
- The draft’s proposed “10 per cent visual disclaimer” is symbolically strong but technically weak.
Way Forward
- Verification infrastructure: Build a digital provenance framework, akin to Aadhaar, for authenticity, where each piece of content carries an invisible but verifiable signature.
- Tiered accountability: Differentiate between platforms that host, generate, or monetise synthetic media. Responsibility should rise with influence.
- AI literacy: Equip citizens to detect manipulation.
| International Case Studies ● The EU’s AI Act mandates watermarking of synthetic content. ● The US relies on voluntary corporate commitments. ● China requires government pre-approval for “deep synthesis” media. |
Conclusion
Democracy runs on trust. And trust is fragile when truth becomes fluid. The solution isn’t censorship. Regulate authenticity, not opinion. If India can institutionalise transparency in AI-generated media, it won’t just protect its elections. It will export a model of digital responsibility for the world.
Source: Indian Express
