Minister for Electronics and Information Technology Ashwini Vaishnaw on Thursday said the government would make rules around deepfakes, synthetic media that mimics authentic images, video and audio, and set out rules to prevent their dissemination online. Mr. Vaishnaw met with representatives from social media and technology firms on the day on the subject, and said the companies were all of the view that action was warranted.
The action follows a furore over the subject after a deepfake of actor Rashmika Mandanna went viral on Instagram and elsewhere earlier this month. Prime Minister Narendra Modi also spoke briefly on the dangers of deepfakes at a Deepavali event. While advisories to social media firms have been sent warning them that the law already prevents impersonation online, the IT Ministry is saying that it will now move swiftly to address the issue more comprehensively.
“Within 10 days, we will come up with clear actionable items on four pillars,” Mr. Vaishnaw said. They are: improving the detection of deepfakes; preventing them from spreading rapidly; improving content reporting mechanisms on social media for synthetic media (timelines currently prescribe a 72-hour limit for removing them, beyond which firms lose legal protections for such content); and building greater awareness on deepfakes among the public.
Mr. Vaishnaw did not specify what legal instrument would be used for this upcoming regulation, saying there had not yet been a decision on whether the rules would come as a subordinate legislation to an existing law, whether a fresh law would be passed, or an existing Act would be amended.
‘More needs to be done’
Labelling or watermarking AI-generated content figured majorly in the discussion with social media platforms, Mr. Vaishnaw said. “This is a basic minimum,” the Minister said of the requirement. “On top of this, more will need to be done.”
“Users have a right to know what is natural and what is not natural,” Mr. Vaishnaw said. “Free speech and privacy are both constructs that are important for us. And these constructs will be undermined by deepfakes.” It is unclear if the draft of the upcoming rules will be published before they are finalised, but Mr. Vaishnaw said the government would conduct an open consultative process beforehand, and take inputs from the public.