The government, he said, would come up with actionable items on four pillars within 10 days.
To combat what IT and Telecom Minister Ashwini Vaishaw called “a threat to democracy,” the government is drafting new regulations that could penalize both the platform and the creator of deepfakes.
New protection regulations are being considered, and they will look at things like watermarking AI-generated content, deepfake detection, rules for data bias, privacy, and guards against concentration. This comes amid reports from some celebrities that their faces were manipulated onto another video.
“A new danger to democracy is the emergence of deepfakes. These (may) erode confidence in society and its institutions,” Vaishnaw stated following a meeting with a number of stakeholders, including academics studying artificial intelligence (AI), social media platforms, and Nasscom.
He declared, “We will begin drafting the regulations today itself and have a separate regulation for deepfakes within a very short timeframe.”
Within 10 days, he said, the government would provide actionable items on four pillars: identifying deepfakes, stopping the spread of such content, bolstering reporting mechanisms, and raising awareness of the problem.
According to him, deepfakes were a concern shared by all the stakeholders in attendance at the meeting. “It was agreed upon by all social media platforms to have sophisticated technology to identify deepfakes.” In two years, there will be over 120 crore internet users in India, up from the current 80 crore. A technology known as “deepfake” uses artificial intelligence (AI) to modify a person’s look, voice, or behavior in ways that can be convincing and difficult to distinguish from real, unaltered content. The need for an AI regulatory framework in the recently passed Digital India law has become more apparent in light of recent deepfakes.
According to Vaishnaw, deceptive promotions or deepfake ads pose a threat to Indian society at the moment.
“The use of social media guarantees that deepfakes can become viral and spread more quickly and significantly without any checks. This is the reason, he told reporters here, that we must act quickly to increase trust in our democracy and society.
When it was discovered that actor Rashmika Mandanna’s face had been featured in an awkward video earlier this month, deepfakes quickly gained notoriety. It was also reported that Katrina Kaif and Kajol were among the other famous people who fell victim to deepfake.
Prime Minister Narendra Modi issued a warning last week regarding the danger that deepfakes pose.
Vaishnaw issued a warning on Saturday that if social media companies don’t take action to stop deepfakes, they risk losing the protection provided by the Information and Technology Act’s “safe harbour” clause. According to the clause, an online platform cannot be responsible for user-shared content on it.
He claimed that deepfake video makers had discovered ways to even get around watermarks and labeling after the meeting with stakeholders on Thursday. “There must therefore be something that manages to escape it.” The following meeting on the topic is scheduled for the first week of December.
The minister stated that within the next ten days, the government would present specific, implementable ideas on four main areas: identifying deepfakes and misinformation; preventing the spread of misinformation; enhancing reporting mechanisms (in-app reporting mechanisms need to be improved); and raising public awareness.
“Every company has expressed concern in the same way. They realized that this is harmful and that it is not free speech; they realized that much stricter regulation was required,” he continued. “Deepfakes can propagate much more quickly without being checked thanks to social media use, and they go viral within minutes of being uploaded.
According to the minister, immediate action is required to preserve our democracy and build social trust. “We must act on this as soon as possible, whether it be through legal, regulatory, or technological means. We need to take a variety of actions.
He responded, “We can bring this in the form of making amendments to the existing rules or we can bring a new set of regulations,” when asked if the current regulation would change or if a new law would be brought.
“We also talked about labeling and watermarking. Everyone agreed that doing this was necessary; it is the bare minimum that everyone must perform,” he remarked. “We will consider the penalties for both the platform creator and the individual who uploaded or created the content when we draft the regulation. We are stating that the government will implement regulations for deepfakes and artificial intelligence (AI)-generated content, as well as for detection, prevention, strengthening the reporting mechanism, raising awareness, and using technologies for these purposes.
Social media companies and platforms pledged to take all necessary precautions to stop the spread of deepfakes until regulations are made. He stated, “They all stated they are making internal progress and they would like to step up the intensity of those steps.”
Declaring that the government values both privacy and free speech, he claimed that deepfake is undermining both of these concepts. “AI-generated content is not harmful to society, according to the new regulation for deep fake.” The minister gave instances of deepfakes and said that a video showing the chief minister of Madhya Pradesh encouraging people to vote for the opposition party went viral during electioneering.
“That was complete falsehood, deepfake, and profound misinformation. In addition, the platforms that serve as the vehicles for the dissemination of this content must accept accountability for the content that is permitted to be published on them. We must make sure that the individuals who produce these are identified and subject to appropriate sanctions.
“Identifying a deepfake is crucial. Differentiating between synthetic and deepfake content is crucial, the speaker continued.
For the latest tech news and reviews, follow Gadgets 360 on X, Facebook, WhatsApp & Telegram For the latest videos , subscribe to our YouTube channel.
