Prime Minister Narendra Modi, speaking at a Diwali Milan program, expressed concern about AI-generated deepfakes, highlighting them as a significant issue. He urged the media to play a role in educating the public about this matter.
On November 17, Prime Minister Narendra Modi addressed the growing problem of deepfakes in India. Deepfake is a type of artificial intelligence (AI) technology that manipulates media, like photos, videos, and audio, to appear highly realistic even though it’s fake. The recent incident involving actress Rashmika Mandanna highlighted the severity of such attacks.
Speaking at the Diwali Milan program in New Delhi, PM Modi emphasized the misuse of AI in creating deepfakes and called on the media to educate the public about this crisis. The issue has prompted some celebrities, like actor Anil Kapoor, to take legal action against unauthorized deepfakes. Rashmika Mandanna also faced a disturbing deepfake video, where her face was added to another woman’s body.
The problem of deepfakes
The Rashmika Mandanna deepfake incident sparked discussions in India about a problem that could potentially escalate. The controversy began with a brief six-second clip of the actor entering a lift, which went viral. However, it was later revealed that the video originally featured Instagram influencer Zara Patel, with Mandanna’s face added through AI.
Union Minister Rajeev Chandrasekhar expressed the government’s commitment to ensuring the safety and trust of internet users, describing deepfakes as a dangerous and damaging form of misinformation that platforms need to address.
Zara Patel, the woman whose video was manipulated, shared her distress on Instagram, expressing concerns about the increased fear for women and girls posting on social media. She urged people to fact-check content on the internet, emphasizing that not everything online is real.
How to spot deepfakes
MIT, known for its dedicated AI and ML research department, has shared valuable tips to help people distinguish between deepfakes and real videos. Here are a few key points:
1. Focus on the Face: DeepFake manipulations often center around facial transformations, so closely observe facial features.
2. Check Blinking Patterns: Genuine videos typically exhibit natural blinking. Note if the person blinks at a normal rate or too frequently.
3. Observe Lip Movements: Some deepfakes involve lip-syncing, so assess if the lip movements appear natural or manipulated.
While the risk of being deepfaked is generally low for most individuals, as sophisticated manipulations demand extensive training data, particularly of personal photos and videos. Without a substantial collection of such data available online, AI models find it challenging to create flawless deepfakes, especially involving lateral facial views.