What happened?

The Ministry of Electronics and IT (MeitY) has advised social media platforms to remove misleading content created by artificial intelligence (AI), such as deepfakes, within 24 hours under IT Rules. This follows the spread of a deepfake video featuring actor Rashmika Mandanna on various social media platforms.

Deepfake is a form of artificial image, audio, or video produced using a specific kind of machine learning known as "deep learning." Verifying its authenticity can be difficult. Moreover, the increased use of AI has raised concerns regarding stronger legislation on privacy, data protection, and cybercrime.

Why does it matter?

Negative applications of deepfake technology

• Gender-based violence and adult content: According to the State of Deepfakes Report 2023, India ranked 6th among the most susceptible nations to deepfake pornography. AI-based tools have made it worryingly easy for laypersons to create realistic deepfakes. Deepfake pornographic content of celebrities has become a million-dollar industry.

  Research on deepfakes by Sensity AI between 2017-2020 revealed that 96% were pornographic and created without women's consent. Recently, AI-generated nude images of teenage girls were circulated in the US and Spain, made by boys using social media photos of the girls. 
  

• Misinformation and propaganda:  AI-generated synthetic media is being used to commit scams, influence elections, and damage reputations, causing a loss of trust in the democratic systems.

  For the 2024 Lok Sabha elections, the Indian government is thinking about using a law that would require WhatsApp to reveal who sent the first message to combat the spread of fake videos. Such measures can also threaten citizen's privacy.

Positive applications of deepfake technology

• Entertainment industry: Deepfakes allow the recreation of the art of long-deceased artists in museums. It facilitates video editing without requiring a reshoot. Additionally, AI-based deepfake technology is being used to create the graphics for video games.

•  Education and health: Deepfake technology could be used to create educational videos that are more engaging and interactive for students. Moreover, a UK-based health charity used this technology to have David Beckham convey an anti-malaria message in nine languages.

What's next?

Following Rashmika's incident, the Minister of State for Electronics and IT stated the government's focus on protecting its citizens online. The updated 2021 IT Rules require social media platforms to remove misinformation within 36 hours of a user or government report. Section 66D of the 2000 IT Act addresses "punishment for cheating using computer resources."

Recently, at the first-ever AI Safety Summit in the UK, major countries, including India, signed a declaration to combat AI risks. However, AI regulation is still emerging, emphasising the need for continuous engagement, investment and knowledge sharing in the field.

Un-charted

Does the victim's gender matter for justice delivery?

This chart is based on a study by a professor at the London School of Economics and Political Science. The research analysed over 4 lakh FIRs in Haryana and uncovered a notable difference in how cases involving male and female complainants are treated. It found that cases registered by men on behalf of females faced fewer challenges than cases where a woman was listed as the primary complainant.

This contrast in conviction rates challenges the common belief that women misuse protective laws like the Domestic Violence Act and rape legislations, which require a reversed burden of proof, i.e. the accused person has to prove that they didn't commit the crime instead of the complainant proving what happened.

The study highlighted that:

• Women's cases have, on average, a lag of over a month between crime occurrence and registration of the case.

• Women's cases spend at least a month longer in the judiciary.

• Women's complaints are casually dismissed as being made impulsively or emotionally, even in constitutional courts.

Past studies have also shown gender bias among police officers. Therefore, the researcher pointed out that more than creating women-focused police stations or fast-track courts, a systemic change in attitudes is needed to address these underlying disparities.

Bonus Section

Think independently using Musk’s First Principles

First principles thinking was popularized by Elon Musk and embraced by philosophers like Aristotle. It's like boiling down a recipe to its essential ingredients and cooking something unique!

People often follow what's already there. For instance, even though ancient Romans used leather bags and wheeled vehicles, it was only in 1970 that someone thought about combining the two, leading to the invention of the rolling suitcase. Before that, people were only tweaking the fabric, or the bag's design. First principles thinking thus allows you to alter the direction of improvement.

Elon Musk changed how rockets are made. Instead of paying an exorbitant price for rockets, he started from scratch and broke down costs step by step. He made cheaper rockets that are reusable, shaking up how space rockets are usually built. Similarly, Airbnb redefined accommodations, moving beyond traditional hotel industry practices to foster peer-to-peer home sharing.

Go beyond the boundary set by the original vision and really think about "What is the functional outcome you are trying to achieve?"

Before you go, check out this video on Deepfakes, where I cover the issue in detail:

SHARE THIS VICHAAR 🙌

If you enjoyed reading my newsletter, please consider sharing it with your loved ones.

Was this forwarded to you? Sign up here.

If you want to give us feedback, Click here.

Read our previous newsletters here