How are deepfakes videos different from Photoshopped images? How to spot them?

The recent controversy pertaining to the viral deepfake video of actor Rashmika Mandanna has triggered concerns about online safety, particularly of women. Union minister Rajeev Chandrasekhar described deepfakes as ‘more dangerous and damaging form of misinformation’

Deepfake is AI creating fake videos by swapping faces or altering voices.(Image by Freepik)

“Under the IT rules notified in April, 2023 – it is a legal obligation for platforms to ensure no misinformation is posted by any user AND ensure that when reported by any user or govt, misinformation is removed in 36 hrs. If platforms do not comply with this, rule 7 will apply and platforms can be taken to court by aggrieved person under provisions of IPC”, the minister had posted on social media platform X.

ALSO READ: Actress’s deepfake raises questions about online gender-based violence

How are deepfakes different from photoshopped images?

Deepfake technology utilises advanced machine learning algorithms to create or manipulate video content, thereby making it appear as if individuals are saying or doing things they never actually did. 

“This is achieved through a process known as generative adversarial networks (GANs), where two machine learning models work in tandem—one to create the forgeries and the other to detect them. This iterative process results in highly convincing fake videos, far surpassing the static nature of photoshopped images”, Kanishk Gaur, cyber Security, data governance and digital technology expert, told Hindustan Times.

ALSO READ: 94% deepfake adult content videos target entertainment industry celebs: Survey

How to spot a deepfake video?

Spotting a deepfake video can be challenging but not impossible. Kanishk Gaur suggests that one should look for inconsistencies such as unnatural blinking patterns, facial distortions, or mismatches between the voice and lip movements. 

“Additionally, there are emerging tools and software that utilize machine learning to detect deepfakes by analyzing these and other subtle cues that may not be immediately apparent to the human eye”, he added.

What should the govt do to address the deepfake menace?

The IT Act 2000 does provide a framework for addressing impersonation and fraud conducted via digital means, the advent of deepfakes calls for more specialized measures. 

According to the cyber security expert, the government should amend existing laws to specifically address the unique challenges posed by deepfakes. The government should support the development of more sophisticated detection tools that can be used by authorities and the public.

Educational initiatives should be undertaken to inform citizens about the nature of deepfakes and how to critically assess digital content. There is a need to work with tech companies and social media platforms to detect and mitigate the spread of deepfakes.

“The government should engage in dialogue with international bodies to establish global norms and policies for combating deepfakes”, Gaur added.

“The incident involving Rashmika Mandana underscores the urgent need for comprehensive strategies to combat the spread of deepfakes. It is a multifaceted issue that requires not just legal intervention but also technological innovation and public education to ensure the safety and integrity of information in the digital space”, Gaur added.

“There is also need for India and world to build competitive tools to detect Deepfakes, these can be both commercials and open source. Govt must bring in strong measures to deal with Deepfakes. As these could impact upcoming Lok Sabha Elections in Feb 2024”, he added.

“Exciting news! Hindustan Times is now on WhatsApp Channels Subscribe today by clicking the link and stay updated with the latest news!” Click here!

Leave a Comment